1

Centennial clonal stability of asexual Daphnia in Greenland lakes despite climate 1

variability 2

3

4

Maison Dane1 5

N. John Anderson2 6

Christopher L. Osburn3 7

John K. Colbourne1 8

Dagmar Frisch1, * 9

10

1 School of Biosciences, University of Birmingham, UK 11

2 Department of Geography, Loughborough University, UK 12

3 Department of Marine, Earth, and Atmospheric Sciences, North Carolina State University, 13

USA 14

* corresponding author 15

16

(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 24, 2020. ; https://doi.org/10.1101/2020.07.22.208553doi: bioRxiv preprint

2

Abstract 17

Climate and environmental condition drive biodiversity at many levels of biological 18

organisation, from populations to ecosystems. Combined with palaeoecological 19

reconstructions, palaeogenetic information on resident populations provides novel insights 20

into evolutionary trajectories and genetic diversity driven by environmental variability. While 21

temporal observations of changing genetic structure are often made of sexual populations, 22

little is known about how environmental change affects the long-term fate of asexual lineages. 23

Here, we provide information on obligately asexual, triploid Daphnia populations from three 24

Arctic lakes in West Greenland through the past 200-300 years to test the impact of a 25

changing environment on the temporal and spatial population genetic structure. The 26

contrasting ecological state of the lakes, specifically regarding salinity and habitat structure 27

may explain the observed lake-specific clonal composition over time. Palaeolimnological 28

reconstructions show considerable environmental fluctuations since 1700 (the end of the 29

Little Ice Age), but the population genetic structure in two lakes was almost unchanged with 30

at most two clones per time period. Their local populations were strongly dominated by a 31

single clone that has persisted for 250-300 years. We discuss three possible explanations for 32

the apparent population genetic stability: (1) the persistent clones are general purpose 33

genotypes that thrive under broad environmental conditions, (2) clonal lineages evolved 34

subtle genotypic differences that are unresolved by microsatellite markers, or (3) epigenetic 35

modifications allow for clonal adaptation to changing environmental conditions. Our results 36

will motivate research into the mechanisms of adaptation in these populations, as well as their 37

evolutionary fate in the light of accelerating climate change in the polar regions. 38

39

(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 24, 2020. ; https://doi.org/10.1101/2020.07.22.208553doi: bioRxiv preprint

3

Introduction 40

Climate and the environment are major drivers of biodiversity in the broad sense. In addition 41

to higher levels of biological organisation (e.g. communities, ecosystems), these 42

environmental drivers affect species at the population level, including their temporal and 43

spatial genetic structure and diversity [1,2]. The responses of population genetic structure and 44

diversity to environmental forcing need to be considered at a range of temporal and spatial 45

scales to fully evaluate how global change processes affect species' distributions and adaptive 46

capacities [3,4]. Long-term, field-based perspectives that span pre- and post-disturbance time 47

periods (typically > 100 years to cover pre-industrial times) are required to balance 48

experimental and laboratory approaches at elucidating ecological responses and genetic 49

adaptation to climate change and chemical pollutants [5]. Studies that combine 50

palaeoecological data with molecular data are especially promising to advance our 51

understanding of past and current adaptation to environmental shifts [3,6]. 52

Palaeolimnological approaches have a long history of providing detailed reconstructions of 53

ecosystem variability at a range of temporal scales (100 to 103 yrs) and across trophic levels 54

[7]. Recent methodological progress combined with a better understanding of preservational 55

constraints and artefacts (e.g. seeds, eggs, cysts) has allowed the use of bulk (environmental) 56

and compound-specific (organismal) DNA to address questions about long-term genetic 57

trends, variability and adaptation in aquatic ecosystems [5,8]. The application of DNA 58

extracted from dormant propagules can also be coupled with resurrection approaches [8,9]. In 59

optimal conditions, the palaeogenetic record allows the reconstruction of temporal patterns of 60

genetic structure and diversity of aquatic populations and communities. These palaeogenetic 61

reconstructions can be coupled with more standard palaeoecological proxies (microfossils, 62

geochemical markers, etc.) that reflect environmental and climatic drivers and forcing. 63

Genetic diversity is often reduced during times of environmental disturbance following strong 64

selection and resulting population bottlenecks [10]. Shifts in genetic structure associated with 65

rapid environmental change have also been observed in long-term studies [11,12]. Likewise, 66

the spatial genetic structure of populations is strongly related to environmental gradients, 67

creating patterns of isolation by environment [13]. Many species have undergone adaptive 68

evolution to shifts in environmental conditions [14]. While these observations are generally 69

made in sexual populations, little is known about long-term dynamics in the genetic make-up 70

of asexual populations. Under an environmental change scenario, asexual lineages may be at a 71

disadvantage compared with sexuals, by shuffling their genetic material thus creating the 72

(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 24, 2020. ; https://doi.org/10.1101/2020.07.22.208553doi: bioRxiv preprint

4

variation needed to fuel selection [15,16]. However, the potential of asexual lineages to 73

persist during the rapid change currently recorded across ecosystems remains unclear. 74

Obligate parthenogenesis in the Daphnia pulex species complex involves the production of 75

apomictic eggs [17,18] that are deposited in an ephippium. This chitinous structure is derived 76

from maternal tissue, and each contains two dormant eggs, which in asexual populations are 77

genetically identical. Ephippia with unhatched eggs can be extracted from the sediment and 78

their DNA examined, allowing for a detailed study of temporal population genetic structure of 79

Daphnia populations [12,19–22]. Long-term studies at the scale of centuries on the fate of 80

asexual Daphnia lineages do not exist. However, various asexual Daphnia lineages have been 81

recorded for 20-30 years in the Canadian Subarctic [23], and for over 70 years in an African 82

lake [20]. Contrasting the idea of centennial persistence of asexual lineages is the age estimate 83

of only decades [24] for asexual lineages of the North American Daphnia pulex-complex that 84

arose by contagious asexuality [25]. 85

Arctic lakes in Greenland provide an ideal system to study long-term dynamics of asexual 86

populations. Many lakes are fishless and inhabited by populations of the keystone herbivore 87

Daphnia, in particular the large-bodied Daphnia pulex-complex [7]. Arctic Daphnia 88

populations are generally obligate asexuals [26,27], and many of these lineages are triploids 89

[28]. In the Illulissat area of West Greenland, clonal richness of Daphnia pulex populations 90

was between 4 and 5 genotypes [29,30]. The lakes along Kangerlussuaq in SW Greenland are 91

particularly well-studied in relationship to recent climate change, with evidence for a dramatic 92

increase in environmental forcing of these sensitive habitats [31]. In the context of the present 93

study, the Kangerlussuaq area is a natural experimental site, in that direct anthropogenic 94

impacts are limited or absent. But there have been substantial environmental changes 95

reflecting climate (temperature, precipitation and altered seasonality) as well as related 96

landscape changes including soil erosion, dust deflation, lake shrinkage, and terrestrial 97

vegetation changes [32]. Kangerlussuaq, in contrast to most Arctic regions, was cooling 98

throughout much of the 20th century. However, since 1996 the area has undergone 99

pronounced climate change, including rapid warming (> 2 °C) over the last 15 years, partly in 100

relation to changes in the Greenland Blocking Index [31]. 101

In order to advance our understanding of the fate of asexual populations in the light of current 102

environmental change at high latitudes, there is a need for empirical data, especially on 103

decadal to centennial timescales. Here, we examine genetic diversity in polymorphic 104

microsatellite DNA, and spatio-temporal genetic structure of three Daphnia lake populations 105

(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 24, 2020. ; https://doi.org/10.1101/2020.07.22.208553doi: bioRxiv preprint

5

in West Greenland near Kangerlussuaq and their relationship with the known regional 106

environmental history. We test the hypothesis that rapid environmental change over the last 107

200-300 years has driven change in temporal and spatial population genetic structure of 108

asexual Daphnia populations at the individual lake and landscape scale. 109

110

Material and Methods 111

Study area 112

The Kangerlussuaq area of SW Greenland is a major lake district with thousands of lakes, 113

with a reasonably well understood regional limnology [33]. This, coupled with preliminary 114

zooplankton surveys and the fact that a number of the lakes have been the subject of 115

palaeolimnological analyses (e.g. [34,35]) meant that sites could be chosen to optimize 116

ephippia recovery, namely deep seasonally anoxic basins. Braya Sø (SS4) was also known to 117

have high concentrations of sediment biomarkers [36]. The area around the head of the fjord 118

(Fig. 1) has low effective precipitation and is characterized by limited hydrological 119

connectivity resulting in a number of closed-basin oligosaline lakes (conductivity range 120

1000–4000 µS cm-1). In general, the Kangerlussuaq freshwater lakes at the head of the fjord 121

have conductivities around 200-500 µS cm-1 and DOC concentrations ca. 30-50 mg L–1; DOC 122

in the oligosaline lakes can be much higher (80-100 mg L-1 [37]; nutrient and major ion water 123

chemistry is given in Table S1). The study lakes are located within a few km of each other 124

(Fig. 1). SS1381 and SS1590 are small (21.5 and 24.6 ha respectively) scour basins with 125

maximum depths around 18 m. Both lakes stratify strongly with anoxic hypolimnia. Braya Sø 126

is a larger, meromictic oligosaline lake (73 ha). The ice-free period is from late-May / early-127

June to late September. The lakes can be considered pristine in comparison to temperate 128

systems in North America and NW Europe and have no direct cultural impact apart from low 129

levels of atmospheric pollution [38]. The catchment vegetation of all lakes is dwarf shrub 130

tundra. 131

The study lakes are presently fishless, which is required for Daphnia populations to persist 132

[39]. All the lakes have variable water levels both at inter-annual and decadal timescales, 133

resulting from their sensitivity to the seasonality of precipitation, the extent of the spring melt 134

input and intense evaporation during the summer [40]. Although the lakes are isolated basins 135

today, SS4 was once part of a group of lakes that formed a large palaeolake in the early 136

Holocene [41]; SS1590 would have drained into this larger lake for a brief period. SS1381 is 137

(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 24, 2020. ; https://doi.org/10.1101/2020.07.22.208553doi: bioRxiv preprint

6

located 6 km to the west of SS1590 and above the level of the highest shoreline of the 138

palaeolake. 139

West Greenland has undergone considerable environmental change (over different timescales) 140

since deglaciation. Research has focussed on a range of biophysical aspects of the ice-land-141

water continuum, including glacier mass balance, terrestrial vegetation and herbivore 142

dynamics, and aquatic biogeochemistry [32]. Lake sediment records have provided details of 143

ecological response to climate and landscape change over centennial and millennial 144

timescales, including diatom, pigment and palaeoclimatic reconstructions [34,35,42,43]. 145

Because ephippia recovery was greatest at SS4 and this lake has a well-defined environmental 146

history, discussion of microsatellite variability and environmental history is restricted to this 147

site. 148

149

Sediment sampling and analysis 150

Short sediment cores (4 replicates per lake) were taken in July 2015 with a 9.5 cm diameter 151

Hon-Kajak sediment corer from the deepest part of each lake. Recovery was around 25-cm at 152

each lake and at all sites included an intact sediment water interface (Fig. S1). Cores were 153

sectioned on shore at 1-cm intervals and samples were placed in individual plastic bags. 154

Although sedimentation rates are low in these lakes (~0.3 cm yr–1; refs), a 1-cm interval was 155

used to maximise recovery of ephippia; finer interval sampling would have provided too few 156

ephippia. Samples were kept refrigerated after return from the field until further analysis. A 157

sub-sample of each core slice was analysed for dry weight and organic matter content using 158

standard methods (loss-on-ignition at 550 °C). The cores used in this study were not dated 159

radiometrically but previous cores have been dated using 210Pb and 14C [43,44]. Organic 160

matter profiles are distinctive at each lake (Fig. S2), allowing among-core correlation and thus 161

chronologies to be transferred from the dated cores to those used in this study. This method is 162

illustrated for SS4 where 4 replicate cores were taken in 2015; core OM profiles are clearly 163

repeatable (Fig. S2). Samples used for microsatellite analysis cover the most recent 200–300 164

years for SS4 and SS1381. The higher sedimentation rate at SS1590 meant that recovery of 165

ephippia with eggs suitable for microsatellite analysis was confined to the most recent ~30 166

years based on 210Pb analysis (NJ Anderson unpublished). 167

Previously, the changing abundance of purple sulphur bacteria (PSB) at Braya Sø has been 168

inferred from the variable concentration of the carotenoid okenone in the lake sediment (see 169

[34]). Unfortunately, pigments were not measured on the sediment cores used in this study 170

(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 24, 2020. ; https://doi.org/10.1101/2020.07.22.208553doi: bioRxiv preprint

7

but we have shown a good agreement between sediment porewater fluorescence of the cores 171

sampled in 2015 and okenone abundance from a 2001 core (Osburn & Anderson 172

unpublished). Therefore, in this study we used the second component (C2) of a PARAFAC 173

model based on the fluorescence excitation emission matrices (EEMs) of porewater 174

chromophoric dissolved organic matter (CDOM) as a proxy for changing PSB abundance. 175

Component 2 exhibits a clear peak centred on 280 nm and is a protein-like fluorescence 176

marker associated with bacteria. Porewater CDOM fluorescence was measured on separate 177

Varian Eclipse spectrofluorometers and fluorescence intensity was calibrated into Raman 178

units. 179

The organic C flux was calculated as in previous studies [45] and at SS4 the ephippia 180

accumulation rate was estimated as the total number of ephippia in a 1-cm sediment slice 181

divided by the linear sediment accumulation rate (yrs cm–1). 182

183

Molecular analysis 184

The studied species was identified as Daphnia pulex sensu lato (in the following: Daphnia 185

pulex), and clustered with lineages of the Polar Daphnia pulicaria clade based on the 186

mitochondrial ND5 [46] (D. Frisch, unpublished data). 187

Ephippia were removed from 1-cm-sections of sediment and enumerated for each sediment 188

section as described in Frisch et al. [12]. Eggs were removed from the ephippia and used for 189

DNA extraction (using one egg per ephippium). DNA was isolated from individual eggs with 190

the QIAamp Microkit (Qiagen Inc, Valencia, CA, USA) following the protocol for DNA 191

extraction from tissue specified in the manufacturer’s manual. To facilitate DNA extraction, 192

eggs were perforated with a sterile micropipette tip, breaking the membrane and exposing egg 193

contents to the extraction buffers. The total number of DNA isolates was 92 from three lakes 194

(n = 57 (SS4), n = 12 (SS1590), n = 23 (SS1381)). 195

Microsatellite loci were amplified in single, 12.5 µl multiplex reactions (Type-it PCR kit, 196

Qiagen Inc, Valencia, CA, USA), using an Eppendorf Nexus Thermal Cycler with thermal 197

cycle conditions recommended in the Type-it PCR kit manual. Ten microsatellite primers 198

(Table S2) representing genome-wide loci were used for genotyping; details in [12,47]). Two 199

primers (Dp90, Dp377) failed to amplify in a consistent manner and were therefore excluded 200

from further analysis. Amplified microsatellites were genotyped on an Applied Biosystems 201

3730 genetic analyser. We used the microsatellite plugin for Geneious 7.0.6 202

(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 24, 2020. ; https://doi.org/10.1101/2020.07.22.208553doi: bioRxiv preprint

8

(https://www.geneious.com) for peak calling and binning. Called peaks were visually 203

inspected and manually adjusted when necessary. 204

All analyses were run in R version 3.6.2 [48] using the R package poppr 2.8.3 [49]. For the 205

population genetic analyses, we excluded all isolates with >5% missing information, yielding 206

a set of 59 isolates (SS4: 28 isolates; SS1381: 10 isolates; SS1590: 21 isolates, Table S3). 207

Analyses included the estimation of allelic diversity for each locus, and the population genetic 208

parameters of estimated multilocus genotypes (eMLG) by rarefaction to the smallest 209

population size of 10 in SS1590. 210

Population genetic structure inferred by the number of clusters of genetically related 211

individuals was analysed using Discriminant Analysis of Principal Components (DAPC) 212

[50].This procedure involves the computation of a Principal Components Analysis (PCA) 213

followed by Discriminant Analysis (DA). The number of PCs to be used in DA was 214

identified by the function xvalDapc() implemented in poppr. 215

216

Results 217

Environmental variability and ephippial dynamics 218

Regional climate in South West Greenland was variable from the end of the Little Ice Age 219

(LIA) with alternating warm and cold periods, and the most recent warm period beginning at 220

the end of the 20th century (Fig. 2A). As example for the in-lake temporal variability in the 221

last centuries, we focus on reconstruction of environmental conditions in SS4. Here, both the 222

OC accumulation rate and the abundance of PSB (as fluorescence PARAFAC component C2) 223

varied over the last 300 years suggesting variable lake production (Fig. 2B). Daphnia pulex 224

ephippia flux varied from ca. 2 to 900 ephippia m2 yr-1 in 1700 to post 1950, peaking at 888 225

around 1900 (Fig. 2C), indicative of considerable fluctuations in population size. 226

227

Population genetic parameters 228

Genetic diversity and population structure of Daphnia populations were studied in three lakes 229

near Kangerlussuaq, SW Greenland (SS4, SS1381, SS1590, Fig. 1), using dormant eggs 230

deposited in the sediment. The covered time period differed between lakes and was 200-300 231

years in SS4 and SS1381, and ~30 years in SS1590. All Daphnia clones were putatively 232

triploid, based on the criterion of three different alleles at a minimum of one locus in each 233

multi-locus genotype (MLG). None of the MLGs included a locus with more than three 234

(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 24, 2020. ; https://doi.org/10.1101/2020.07.22.208553doi: bioRxiv preprint

9

alleles. Of the eight loci used for genotyping, six (Dp162, Dp291, Dp369, Dp401, Dp437, 235

Dp461) had three different alleles in at least one individual, while two (Dp43, Dp173) had a 236

maximum of two alleles. 237

Clonal diversity differed between the three lakes (Table 1): of 28 eggs examined in SS4, we 238

found a total of three MLGs (maximum of 2 per time period, Fig. 3A) of which one was 239

dominant throughout three centuries (MLG1). Only a single clone (MLG10) was detected in 240

the 21 isolates from SS1381, which persisted over several centuries, as did MLG1 in SS4. In 241

contrast, the Daphnia population in SS1590 was more diverse than the former two lakes with 242

seven MLGs in a total of 10 isolates, of which only one MLG was found during two (non-243

adjacent) time periods. These differences in clonal diversity and dominance structure are also 244

reflected in the rarefied estimate of clonal diversity (eMLG, Table 1), which in SS1590 was 245

about 3.5 and 6 times higher than in SS4 and SS1381, respectively. No clear temporal pattern 246

was discernible in either of the two lakes with long-term records (Fig. 3A, SS4 and SS1381). 247

The Daphnia lineages of the three lake populations formed well-defined clusters (DAPC, Fig. 248

3B), separating the lakes at almost equal distances. The analysis also illustrates the genetic 249

similarity of three MLGs of SS4. While the clones of SS1590 were more scattered they still 250

formed their own cluster without overlap to the other lakes. 251

252

Discussion 253

The results of this study indicated almost equal separation of the three Daphnia populations 254

among the lakes (Fig. 3) and despite considerable in-lake and regional environmental change 255

since ~1700 AD (discussed below), the clonal structure of Daphnia populations in two of 256

three lakes (SS4 and SS1381) was remarkably stable. Importantly, we found strong 257

dominance of a single, lake-specific asexual clone throughout the past two to three centuries 258

in each of the two lakes. 259

260

Spatial patterns 261

The analysis of population genetic structure revealed almost equal separation of the three 262

populations (Fig. 3), a spatial pattern that may reflect the primary environmental differences 263

among the lakes as well as the historic connectivity between these lakes: SS4 is oligosaline 264

and thus differs strongly from SS1381 and SS1590. The latter two are quite similar to each 265

(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 24, 2020. ; https://doi.org/10.1101/2020.07.22.208553doi: bioRxiv preprint

10

other in terms of their water chemistry (Table S1) and typical of many of the freshwater lakes 266

around the head of the fjord [33]. Clearly, ionic composition and inorganic C chemistry is an 267

important driver of genotypic variability as much as it determines regional biodiversity and 268

community structure (diatoms, chrysophytes; [51,52]. Conductivity is the dominant control on 269

algal composition in this area and is highly correlated with both DOC concentration and 270

CDOM quality [37]. Conductivity was also identified as one of the major factors driving 271

clonal composition of Daphnia populations in an area about 250 km north of Kangerlussuaq 272

[30]. Lakes SS1381 and SS1590 have never been connected hydrologically, since SS1381 lies 273

outside the boundary of the large palaeolake that included SS4 in the period immediately after 274

deglaciation [41]. The lack of shared clones between lakes suggests local adaptation to the 275

environmental conditions of each resident lake, in particular given the geographic proximity 276

between all three study lakes (less than 10 km) and a likely lack of dispersal limitation as 277

observed in other populations of West Greenland Daphnia [29]. 278

The overall greater clonal diversity of the SS1590 Daphnia population compared with that of 279

SS4 and SS1381 may reflect the greater diversity of habitats in this lake – its two sub-basins 280

are quite different in terms of their morphometry and substrate variability, including the 281

dominant macrophytes (Anderson, unpublished field observations). Not only are littoral and 282

benthic habitats more diverse in SS1590, but the lake also has a more dynamic response to 283

evaporative driven lake level changes than many of the lakes in the area. These short-term 284

lake level changes also exacerbate habitat heterogeneity more at this site than others, due to 285

its morphometry. SS1381 has a simpler morphometry, essentially with one deep basin. SS4, 286

the largest of the lakes can be considered a pelagic system and while lake levels are variable 287

here as well, the contribution of the littoral zone to whole lake production and diversity is 288

probably more limited. 289

290

Regional environmental change 291

Lakes in the Kangerlussuaq area are tightly coupled to regional climate change which 292

influences both terrestrial landscape and in-lake aquatic processes, such as hydrological 293

runoff, terrestrial productivity, lake levels, conductivity and DOC concentration [53]. The 294

period covered by the genetic analyses, ~300 years, includes the end of the Little Ice Age 295

(LIA); a period of considerable change in regional climate, with aridity and fluctuating 296

temperatures. Lake levels would have been lower than present [41,43] and aridity affects dust 297

deflation from the sandurs immediately north and to the east of the study area. As the lakes 298

(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 24, 2020. ; https://doi.org/10.1101/2020.07.22.208553doi: bioRxiv preprint

11

are strongly P-limited [54], dust is a possible important nutrient source. Atmospheric reactive 299

N deposition has increased across the Arctic [55] and Greenland is no exception. The 300

changing nutrient balance associated with these varying sources will have impacted primary 301

production and thus secondary producers (i.e. Daphnia). But as well as these broader, 302

regional responses which can be traced across lakes, each individual lake has its own 303

ecological trajectory in environmental space (see for example [35]). Long-term air 304

temperature records show considerable variability during the 20th century, including a period 305

of pronounced cooling. Rapid warming characterizes the start of the 21st century. Lake 306

temperatures track air temperatures tightly in this area [56] and so epilimnetic temperatures 307

would have varied similarly. 308

As an example for local environmental variability since LIA, SS4 showed pronounced 309

variation of okenone, a pigment which can be used as a proxy for purple sulphur bacteria 310

(PSB) abundance [34]. In SS4, in conjunction with indicators of primary production, variation 311

of ephippial densities provides evidence of historic changes in trophic interactions. Ephippial 312

production serves as a rough estimate of Daphnia population size [57,58]. There is some 313

suggestion that over the last ~800 years the Daphnia population in SS4 has been tracking 314

purple sulphur bacteria (Frisch & Osburn unpublished), which Daphnia may exploit as a 315

direct or indirect food source, as observed in other studies [59,60]. 316

317

Long-term persistence of dominant clones 318

Despite these profound environmental changes, there is limited genotypic variability over 319

time. We did not find evidence for environmental dynamics to drive patterns of population 320

genetic structure or genetic diversity over the past several centuries. Our data suggest the 321

presence of a single clone in two of the study lakes, in SS4 (MLG1) and SS1381 (MLG10), 322

across all sample depths, over the past 200 to 300 years. In SS4, two other clones were 323

detected in two different time periods, both with very low abundance. In contrast, the third 324

lake (SS1590) had a higher number of clones. While it is obvious that clonal diversity in this 325

lake overall was much higher, the temporal pattern of genetic structure and diversity cannot 326

be identified with certainty due to the low numbers of isolates; overall ephippial density in 327

this lake was much lower (personal observation) with a smaller amount of well-preserved 328

eggs suitable for genetic analyses. 329

330

(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 24, 2020. ; https://doi.org/10.1101/2020.07.22.208553doi: bioRxiv preprint

12

There are several tenable explanations for the apparent dominance and persistence of single 331

clones: 332

(1) The persistent clone may be a general-purpose genotype that successfully exploits a 333

variety of historic environments that are not (yet) within a harmful range 334

The invasion and dominance of a single, asexual Daphnia clone has been observed in several 335

lakes throughout a wide geographic range. For example, an asexual Daphnia pulex clone 336

invaded the population in Lake Naivasha, Kenia in the 1920s, and then displaced the local 337

sexual, genetically diverse Daphnia population [20]. Successful invasion of the same clone 338

was observed throughout Africa in a wide range of aquatic habitats and environmental 339

conditions, testifying to the exceptional niche breadth of this asexual clone. Similar 340

observations have been made in Japan [61], where four asexual Daphnia clones invaded a 341

large number of aquatic habitats across a wide geographical range and ecological conditions. 342

In our study of Arctic populations, the persistence of Daphnia clones despite major 343

temperature changes over the last centuries in the study area may reflect the considerable in-344

lake temperature gradient that Daphnia experience on a regular basis. At SS4, daily vertical 345

migration to graze directly or indirectly on PSB and POC (DOC) in the metalimnion at SS4 (a 346

distance of some 8-12 m) would expose animals to a temperature change of >10 °C, 347

considerably more than the temperature change during the recent millennia. Moreover, the 348

annual temperature range in the epilimnion is also in the order of 10 °C [44]. Although only 349

SS4 has a metalimnetic PSB plate, all three study lakes have substantial gradients in DOC and 350

POC, suggesting that they are utilising a microbial loop in the lakes. The three study lakes 351

are seasonally anoxic with hypolimnetic reductions in O2 during both summer and winter. It is 352

possible that these seasonal environmental changes and the vertical O2 gradients exert greater 353

physiological stress in Daphnia than the stress associated with regional warming. Of course, 354

this genetic stasis may not continue should environmental change in SW Greenland stay at a 355

similar rate [31]. 356

(2) Undetected by the applied genetic resolution, a single MLG may comprise several distinct 357

clonal lineages, each adapted to a specific historic environment 358

The possibility of cryptic genetic variation must be considered due to the limited resolution 359

offered by microsatellite markers which cannot fully account for possibly existing genome-360

wide variation. For example, using 12 microsatellite loci, So et al. [61] detected only four 361

MLGs in asexual Daphnia pulex that had invaded Japanese lakes and ponds. Interestingly, 362

these MLGs comprised 21 mitochondrial haplotypes, indicating a higher genotype diversity 363

(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 24, 2020. ; https://doi.org/10.1101/2020.07.22.208553doi: bioRxiv preprint

13

than that resolved by the microsatellite loci [61], a finding that was later confirmed using 364

whole genome sequencing [62]. In particular, these authors concluded that the observed 365

divergent traits in these closely related asexual genotypes evolved without recombination 366

from an ancestral clone by a limited number of functionally significant mutations [62]. 367

(3) Epigenetic modifications explain clonal adaptation and allow clones to persist 368

There is increasing evidence that epigenetic mechanisms contribute to evolution [63,64] and 369

to adaptive responses related to climate and environmental change [65–67]. In the absence of 370

genomic variation, it is becoming increasingly evident that changes in epigenetic profiles 371

could allow for rapid, heritable adaptations to environmental cues that precede the more 372

slowly evolving changes in DNA sequences [68], i.e. the classical Darwinian mutation 373

accumulation and selection concept. Epigenetic modifications would allow persistence of 374

asexual clones in the absence of recombination and generation of genetic variation, because 375

transgenerational inheritance of DNA methylation is highly likely in Daphnia, whose gametes 376

are derived from almost fully matured tissue [69]. 377

378

Synthesis 379

In order to test the ecological implications of these findings, laboratory experiments are 380

needed in future studies. In particular, representatives of the current populations can be used 381

to test whether the dominant clones exhibit phenotypic plasticity with a wide tolerance 382

towards local conditions (temperature, food, salinity), and whether local adaptation to these 383

conditions can be observed. If possible, hatchlings of eggs from older sediments should be 384

included to compare phenotypic and transcriptomic responses of ancient and contemporary 385

isolates [12,70]. In addition, the role of epigenetic modifications should be considered in 386

further studies. 387

These results, to our best knowledge, are the first to report the population genetic structure of 388

asexual, polyploid Daphnia populations of the circumpolar Daphnia pulex-complex over 389

century timescales. Future, more intensive sampling may reveal additional clonal lineages; 390

however, the observed dominance patterns and persistence of individual clonal lineages are 391

unlikely to change. Our results stimulate several questions relating to the mechanisms of 392

adaptation in these populations, as well as their evolutionary fate during the next decades and 393

beyond, on the assumption that climate change in Greenland and other Arctic regions 394

proceeds on the predicted trajectory with severe ecological consequences. 395

(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 24, 2020. ; https://doi.org/10.1101/2020.07.22.208553doi: bioRxiv preprint

14

396

Acknowledgements 397

We are grateful to Erika Whiteford, Madeleine Giles, Amanda Burson and Tania Cresswell‐398

Maynard for help with field work, to Chris O'Grady for assistance in the laboratory, to Andy 399

Moss for providing laboratory space for core processing at UoB and to Fengjuan Xiao 400

(Lboro) for the loss-on-ignition analyses. DF received funding from the European Union’s 401

Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant 402

agreement No. 658714. NJA’s work in Kangerlussuaq is the supported by NERC. 403

404

Supplementary Information 405

Fig. S1 Photograph of a sediment core recovered in SS4 406

Fig. S2. A-B. Correlation of replicate cores at SS4 C. Core correlation at SS1381 407

Table S1 Nutrient and major ion water chemistry in the three study lakes 408

Table S2 Details on allelic diversity for eight microsatellite loci of the three Daphnia pulex 409

s.l. populations in this study 410

(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 24, 2020. ; https://doi.org/10.1101/2020.07.22.208553doi: bioRxiv preprint

15

Table 1. Genetic diversity of the Daphnia pulex population in three lakes of West Greenland 411

integrated over several sediment depths. 412

413

414

415 Pop N MLG eMLG SE

SS4 28 3 1.7 0.7 SS1590 10 6 6.0 0.0

SS1381 21 1 1.0 0.0 Total 59 10 3.9 1.1

(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 24, 2020. ; https://doi.org/10.1101/2020.07.22.208553doi: bioRxiv preprint

16

Fig. 1 Overview of the study area. Insert shows location of the study lakes SS4 (Braya Sø), 416

SS1590, and SS1381. 417

418

419

(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 24, 2020. ; https://doi.org/10.1101/2020.07.22.208553doi: bioRxiv preprint

17

Fig. 2 Environmental history at SS4 (Braya Sø) and the response of Daphnia genotypes. A. 420

Climate variability: reconstructed summer lake water temperatures based on alkenones at 421

Braya Sø and Lake E (see [44] for details) and the mean annual air temperature recorded at 422

Nuuk. B. Aquatic production indices for the lake: organic C burial reflects total in-lake 423

production and abundance of purple sulphur bacteria inferred from porewater fluorescence (as 424

Parafac Component C2). C. The stability of the dominant genotype over time period covered 425

by the sediment analyses is indicated by the solid red line; two secondary, minor genotypes 426

(GT) are indicated (solid orange and dotted red lines). 427

428

429

(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 24, 2020. ; https://doi.org/10.1101/2020.07.22.208553doi: bioRxiv preprint

18

Fig. 3 Spatial and temporal population genetic structure of the Daphnia populations in three 430

lakes. To avoid an artificial increase of identical genotypes, only one egg per ephippium was 431

used for microsatellite analysis. A. Abundance of multilocus genotypes (MLG1 to MLG10) 432

with information on lake and sediment age from which eggs were isolated. Each row 433

represents a lake: SS4 (4 time periods), SS1381 (4 time periods), and SS1590 (three time 434

periods). B. DAPC of MLGs identified in the three lakes and time periods. Fill colours 435

represent MLGs and correspond to colours in A. 436

437

(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 24, 2020. ; https://doi.org/10.1101/2020.07.22.208553doi: bioRxiv preprint

19

References 438

1. Pauls SU, Nowak C, Bálint M, Pfenninger M. 2013 The impact of global climate 439 change on genetic diversity within populations and species. Mol. Ecol. 22, 925–946. 440 (doi:10.1111/mec.12152) 441

2. Scheffers BR et al. 2016 The broad footprint of climate change from genes to biomes 442 to people. Science 354, aaf7671. (doi:10.1126/science.aaf7671) 443

3. Ellegaard M et al. 2020 Dead or alive: sediment DNA archives as tools for tracking 444 aquatic evolution and adaptation. Commun. Biol. 3. (doi:10.1038/s42003-020-0899-z) 445

4. Orsini L, Schwenk K, De Meester L, Colbourne JK, Pfrender ME, Weider LJ. 2013 446 The evolutionary time machine: forecasting how populations can adapt to changing 447 environments using dormant propagules. Trends Ecol. Evol. 28, 274–282. 448

5. Nogués-Bravo D, Rodríguez-Sánchez F, Orsini L, de Boer E, Jansson R, Morlon H, 449 Fordham DA, Jackson ST. 2018 Cracking the code of biodiversity responses to past 450 climate change. Trends Ecol. Evol. 33, 765–776. (doi:10.1016/j.tree.2018.07.005) 451

6. Napier JD, de Lafontaine G, Chipman ML. 2020 The evolution of paleoecology. 452 Trends Ecol. Evol. 35, 293–295. (doi:10.1016/j.tree.2019.12.006) 453

7. Jeppesen E, Lauridsen TL, Christoffersen KS, Landkildehus F, Geertz-Hansen P, 454 Amsinck SL, Søndergaard M, Davidson TA, Rigét F. 2017 The structuring role of fish 455 in Greenland lakes: an overview based on contemporary and paleoecological studies of 456 87 lakes from the low and the high Arctic. Hydrobiologia 800, 99–113. 457 (doi:10.1007/s10750-017-3279-z) 458

8. Weider LJ, Jeyasingh PD, Frisch D. 2018 Evolutionary aspects of resurrection ecology: 459 Progress, scope, and applications-An overview. Evol. Appl. 11, 3–10. 460

9. Burge DRL, Edlund MB, Frisch D. 2018 Paleolimnology and resurrection ecology: 461 The future of reconstructing the past. Evol. Appl. 11, 42–59. (doi:10.1111/eva.12556) 462

10. Banks SC, Cary GJ, Smith AL, Davies ID, Driscoll DA, Gill AM, Lindenmayer DB, 463 Peakall R. 2013 How does ecological disturbance influence genetic diversity? Trends 464 Ecol. Evol. 28, 670–679. (doi:10.1016/j.tree.2013.08.005) 465

11. Lundholm N, Ribeiro S, Godhe A, Rostgaard Nielsen L, Ellegaard M. 2017 Exploring 466 the impact of multidecadal environmental changes on the population genetic structure 467 of a marine primary producer. Ecol. Evol. 7, 3132–3142. (doi:10.1002/ece3.2906) 468

12. Frisch D, Morton PK, Roy Chowdhury P, Culver BW, Colbourne JK, Weider LJ, 469 Jeyasingh PD. 2014 A millennial-scale chronicle of evolutionary responses to cultural 470 eutrophication in Daphnia. Ecol. Lett. 17, 360–368. 471

13. Wang IJ, Bradburd GS. 2014 Isolation by environment. Mol. Ecol. 23, 5649–5662. 472 (doi:10.1111/mec.12938) 473

14. Carroll SP, Hendry AP, Reznick DN, Fox CW. 2007 Evolution on ecological time-474 scales. Funct. Ecol. 21, 387–393. (doi:10.1111/j.1365-2435.2007.01289.x) 475

15. Crow JF. 1994 Advantages of sexual reproduction. Dev. Genet. 15, 205–213. 476 (doi:10.1002/dvg.1020150303) 477

16. Lachapelle J, Bell G. 2012 Evolutionary rescue of sexual and asexual populations in a 478

(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 24, 2020. ; https://doi.org/10.1101/2020.07.22.208553doi: bioRxiv preprint

20

deteriorating environment. Evolution 66, 3508–3518. (doi:10.1111/j.1558-479 5646.2012.01697.x) 480

17. Dufresne F, Marková S, Vergilino R, Ventura M, Kotlík P. 2011 Diversity in the 481 reproductive modes of European Daphnia pulicaria deviates from the geographical 482 parthenogenesis. PLoS One 6, e20049. http://doi.org/10.1371/journal.pone.002004. 483

18. Innes DJ, Dufresne F. 2020 Parthenogenesis. In The Natural History of the Crustacea: 484 Reproductive Biology (eds R Cothran, M Thiel), p. 432. Oxford University Press. 485

19. Limburg PA, Weider LJ. 2002 ‘Ancient’ DNA in the resting egg bank of a 486 microcrustacean can serve as a palaeolimnological database. Proc. R. Soc. London Ser. 487 B-Biological Sci. 269, 281–287. 488

20. Mergeay J, Verschuren D, De Meester L. 2006 Invasion of an asexual American water 489 flea clone throughout Africa and rapid displacement of a native sibling species. Proc. 490 R. Soc. B-Biological Sci. 273, 2839–2844. 491

21. Orsini L, Marshall H, Cuenca Cambronero M, Chaturvedi A, Thomas KW, Pfrender 492 ME, Spanier KI, De Meester L. 2016 Temporal genetic stability in natural populations 493 of the waterflea Daphnia magna in response to strong selection pressure. Mol. Ecol. 494 25, 6024–6038. (doi:doi:10.1111/mec.13907) 495

22. Frisch D, Morton PK, Culver BW, Edlund MB, Jeyasingh PD, Weider LJ. 2016 496 Paleogenetic records of Daphnia pulicaria in North American lakes reveal the impact 497 of cultural eutrophication. Glob. Chang. Biol. 23, 708–718. (doi:doi: 498 10.1111/gcb.13445) 499

23. Weider LJ, Frisch D, Hebert PDN. 2010 Long-term changes in metapopulation genetic 500 structure: a quarter-century retrospective study on low-Arctic rock pool Daphnia. Proc. 501 R. Soc. B-Biological Sci. 277, 139–146. 502

24. Tucker AE, Ackerman MS, Eads BD, Xu S, Lynch M. 2013 Population-genomic 503 insights into the evolutionary origin and fate of obligately asexual Daphnia pulex. 504 Proc. Natl. Acad. Sci. 110, 15740–15745. (doi:10.1073/pnas.1313388110) 505

25. Paland S, Colbourn JK, Lynch M. 2005 Evolutionary history of contagious asexuality 506 in Daphnia pulex. Evolution 59, 800–813. 507

26. Weider LJ, Hobaek A, Crease TJ, Stibor H. 1996 Molecular characterization of clonal 508 population structure and biogeography of arctic apomictic Daphnia from Greenland 509 and Iceland. Mol. Ecol. 5, 107–118. 510

27. Decaestecker E, de Meester L, Mergeay J. 2009 Cyclical parthenogenesis in Daphnia: 511 sexual versus asexual reproduction. In Lost Sex: The Evolutionary Biology of 512 Parthenogenesis (eds I Schön, K Martens, P van Dijk), pp. 295–316. Springer 513 Science+Business Media B.V. 514

28. Vergilino R, Belzile C, Dufresne F. 2009 Genome size evolution and polyploidy in the 515 Daphnia pulex complex (Cladocera: Daphniidae). Biol. J. Linn. Soc. 97, 68–79. 516 (doi:doi.org/10.1111/j.1095-8312.2008.01185.x) 517

29. Haileselasie TH, Mergeay J, Weider LJ, Jeppesen E, De Meester L. 2016 Colonization 518 history and clonal richness of asexual Daphnia in periglacial habitats of contrasting age 519 in West Greenland. J. Anim. Ecol. 85, 1108–1117. (doi:10.1111/1365-2656.12513) 520

(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 24, 2020. ; https://doi.org/10.1101/2020.07.22.208553doi: bioRxiv preprint

21

30. Haileselasie TH et al. 2016 Environment not dispersal limitation drives clonal 521 composition of Arctic Daphnia in a recently deglaciated area. Mol. Ecol. 25, 5830–522 5842. (doi:10.1111/mec.13843) 523

31. Saros JE et al. 2019 Arctic climate shifts drive rapid ecosystem responses across the 524 West Greenland landscape. Environ. Res. Lett. 14, 074027. (doi:10.1088/1748-525 9326/ab2928) 526

32. Anderson NJ et al. 2017 The Arctic in the twenty-first century: changing 527 biogeochemical linkages across a paraglacial landscape of Greenland. Bioscience 67, 528 118–133. (doi:10.1093/biosci/biw158) 529

33. Whiteford EJ, McGowan S, Barry CD, Anderson NJ. 2016 Seasonal and regional 530 controls of phytoplankton production along a climate gradient in South-West 531 Greenland during ice-cover and ice-free conditions. Arctic, Antarct. Alp. Res. 48, 139–532 159. (doi:10.1657/AAAR0015-003) 533

34. McGowan S, Juhler RK, Anderson NJ. 2008 Autotrophic response to lake age, 534 conductivity and temperature in two West Greenland lakes. J. Paleolimnol. 39, 301–535 317. (doi:10.1007/s10933-007-9105-2) 536

35. Law AC, Anderson NJ, McGowan S. 2015 Spatial and temporal variability of lake 537 ontogeny in south-western Greenland. Quat. Sci. Rev. 126, 1–16. 538 (doi:10.1016/j.quascirev.2015.08.005) 539

36. D’Andrea WJ, Lage M, Martiny JBH, Laatsch AD, Amaral-Zettler LA, Sogin ML, 540 Huang Y. 2006 Alkenone producers inferred from well-preserved 18S rDNA in 541 Greenland lake sediments. J. Geophys. Res. 111, G03013. 542 (doi:10.1029/2005JG000121) 543

37. Anderson NJ, Stedmon CA. 2007 The effect of evapoconcentration on dissolved 544 organic carbon concentration and quality in lakes of SW Greenland. Freshw. Biol. 52, 545 280–289. (doi:10.1111/j.1365-2427.2006.01688.x) 546

38. Bindler R, Renberg I, John Anderson N, Appleby PG, Emteryd O, Boyle J. 2001 Pb 547 isotope ratios of lake sediments in West Greenland: inferences on pollution sources. 548 Atmos. Environ. 35, 4675–4685. (doi:10.1016/S1352-2310(01)00115-7) 549

39. Lauridsen T, Jeppesen E, Landkildehus F, Søndergaard M. 2001 Horizontal 550 distribution of cladocerans in arctic Greenland lakes-impact of macrophytes and fish. 551 Hydrobiologia 442, 107–116. (doi:10.1023/A:1017517919996) 552

40. Anderson NJ, Harriman R, Ryves DB, Patrick ST. 2001 Dominant factors controlling 553 variability in the ionic composition of West Greenland lakes. Arctic, Antarct. Alp. Res. 554 33, 418. (doi:10.2307/1552551) 555

41. Aebly FA, Fritz SC. 2009 Palaeohydrology of Kangerlussuaq (Søndre Strømfjord), 556 West Greenland during the last ~8000 years. The Holocene 19, 91–104. 557 (doi:10.1177/0959683608096601) 558

42. Presthus Heggen M, Birks HH, Anderson NJ. 2010 Long-term ecosystem dynamics of 559 a small lake and its catchment in west Greenland. The Holocene 20, 1207–1222. 560 (doi:10.1177/0959683610371995) 561

43. McGowan S, Ryves DB, Anderson NJ. 2003 Holocene records of effective 562 precipitation in West Greenland. Holocene (doi:10.1191/0959683603hl610rp) 563

(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 24, 2020. ; https://doi.org/10.1101/2020.07.22.208553doi: bioRxiv preprint

22

44. D’Andrea WJ, Huang Y, Fritz SC, Anderson NJ. 2011 Abrupt Holocene climate 564 change as an important factor for human migration in West Greenland. Proc. Natl. 565 Acad. Sci. 108, 9765–9769. (doi:10.1073/pnas.1101708108) 566

45. Anderson NJ, Leng MJ, Osburn CL, Fritz SC, Law AC, McGowan S. 2018 A 567 landscape perspective of Holocene organic carbon cycling in coastal SW Greenland 568 lake-catchments. Quat. Sci. Rev. 202, 98–108. (doi:10.1016/j.quascirev.2018.09.006) 569

46. Colbourne JK, Crease TJ, Weider LJ, Hebert PDN, Hobæk A, Dufresne F. 1998 570 Phylogenetics and evolution of a circumarctic species complex (Cladocera: Daphnia 571 pulex). Biol. J. Linn. Soc. 65, 347–365. 572

47. Colbourne JK, Robison B, Bogart K, Lynch M. 2004 Five hundred and twenty-eight 573 microsatellite markers for ecological genomic investigations using Daphnia. Mol. Ecol. 574 Notes 4, 485–490. 575

48. R Core Team (2019). 2019 R: A language and environment for statistical computing. 576 Accessed 1st April 2019 577

49. Kamvar ZN, Tabima JF, Grünwald NJ. 2014 Poppr : an R package for genetic analysis 578 of populations with clonal, partially clonal, and/or sexual reproduction. PeerJ 2, e281. 579 (doi:10.7717/peerj.281) 580

50. Jombart T, Devillard S, Balloux F. 2010 Discriminant analysis of principal 581 components: a new method for the analysis of genetically structured populations. BMC 582 Genet. 11, 94. (doi:10.1186/1471-2156-11-94) 583

51. Ryves DB, McGowan S, Anderson NJ. 2002 Development and evaluation of a diatom-584 conductivity model from lakes in West Greenland. Freshw. Biol. 47, 995–1014. 585 (doi:10.1046/j.1365-2427.2002.00832.x) 586

52. Pla S, Anderson NJ. 2005 Environmental factors correlated with chrysophyte cyst 587 assemblages in low Arctic lakes of Southwest Greenland. J. Phycol. 41, 957–974. 588 (doi:10.1111/j.1529-8817.2005.00131.x) 589

53. Osburn CL, Anderson NJ, Leng MJ, Barry CD, Whiteford EJ. 2019 Stable isotopes 590 reveal independent carbon pools across an Arctic hydro‐climatic gradient: Implications 591 for the fate of carbon in warmer and drier conditions. Limnol. Oceanogr. Lett. 4, 205–592 213. (doi:10.1002/lol2.10119) 593

54. Brutemark A, Rengefors K, Anderson NJ. 2006 An experimental investigation of 594 phytoplankton nutrient limitation in two contrasting low arctic lakes. Polar Biol. 29, 595 487–494. (doi:10.1007/s00300-005-0079-0) 596

55. Wolfe AP et al. 2013 Stratigraphic expressions of the Holocene-Anthropocene 597 transition revealed in sediments from remote lakes. Earth-Science Rev. 598 (doi:10.1016/j.earscirev.2012.11.001) 599

56. Kettle H, Thompson R, Anderson NJ, Livingstone DM. 2004 Empirical modeling of 600 summer lake surface temperatures in southwest Greenland. Limnol. Oceanogr. 49, 601 271–282. (doi:10.4319/lo.2004.49.1.0271) 602

57. Jankowski T, Straile D. 2003 A comparison of egg-bank and long-term plankton 603 dynamics of two Daphnia species, D. hyalina and D. galeata : Potentials and limits of 604 reconstruction. Limnol. Oceanogr. 48, 1948–1955. (doi:10.4319/lo.2003.48.5.1948) 605

(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 24, 2020. ; https://doi.org/10.1101/2020.07.22.208553doi: bioRxiv preprint

23

58. Nykänen M, Vakkilainen K, Liukkonen M, Kairesalo T. 2009 Cladoceran remains in 606 lake sediments: a comparison between plankton counts and sediment records. J. 607 Paleolimnol. 42, 551–570. (doi:10.1007/s10933-008-9304-5) 608

59. Massana R, Gasol JM, Jürgens K, Pedròs-Aliò C. 1994 Impact of Daphnia pulex on a 609 metalimnetic microbial community. J. Plankton Res. 16, 1379–1399. 610 (doi:10.1093/plankt/16.10.1379) 611

60. Jürgens K, Wickham SA, Rothaupt KO, Santer B. 1996 Feeding rates of macro- and 612 microzooplankton on heterotrophic nanoflagellates. Limnol. Oceanogr. 41, 1833–1839. 613

61. So M, Ohtsuki H, Makino W, Ishida S, Kumagai H, Yamaki KG, Urabe J. 2015 614 Invasion and molecular evolution of Daphnia pulex in Japan. Limnol. Oceanogr. 60, 615 1129–1138. (doi:10.1002/lno.10087) 616

62. Tian X, Ohtsuki H, Urabe J. 2019 Evolution of asexual Daphnia pulex in Japan: 617 variations and covariations of the digestive, morphological and life history traits. BMC 618 Evol. Biol. 19, 122. (doi:10.1186/s12862-019-1453-9) 619

63. Becker C, Hagmann J, Müller J, Koenig D, Stegle O, Borgwardt K, Weigel D. 2011 620 Spontaneous epigenetic variation in the Arabidopsis thaliana methylome. Nature 480, 621 245–249. (doi:10.1038/nature10555) 622

64. Asselman J, De Coninck DIM, Pfrender ME, De Schamphelaere KAC. 2016 Gene 623 body methylation patterns in Daphnia are associated with gene family size. Genome 624 Biol. Evol. 8, 1185–1196. (doi:10.1093/gbe/evw069) 625

65. Alakärppä E, Salo HM, Valledor L, Cañal MJ, Häggman H, Vuosku J. 2018 Natural 626 variation of DNA methylation and gene expression may determine local adaptations of 627 Scots pine populations. J. Exp. Bot. 69, 5293–5305. (doi:10.1093/jxb/ery292) 628

66. Metzger DCH, Schulte PM. 2018 The DNA methylation landscape of stickleback 629 reveals patterns of sex chromosome evolution and effects of environmental salinity. 630 Genome Biol. Evol. 10, :775-785. (doi:10.1093/gbe/evy034) 631

67. Münzbergová Z, Latzel V, Šurinová M, Hadincová V. 2019 DNA methylation as a 632 possible mechanism affecting ability of natural populations to adapt to changing 633 climate. Oikos 128, 124–134. (doi:10.1111/oik.05591) 634

68. E. Deakin J, Domaschenz R, Siew Lim P, Ezaz T, Rao S. 2014 Comparative 635 epigenomics: an emerging field with breakthrough potential to understand evolution of 636 epigenetic regulation. AIMS Genet. 1, 34–54. (doi:10.3934/genet.2014.1.34) 637

69. Wojewodzic MW, Beaton MJ. 2017 The future of environmental epigenetics: Insights 638 using the clonal water flea model. Adv. In Insect Phys. 53, 287–312. 639 (doi:10.1016/bs.aiip.2017.05.002) 640

70. Frisch D, Becker D, Wojewodzic MW. 2020 Dissecting the transcriptomic basis of 641 phenotypic evolution in an aquatic keystone grazer. Mol. Biol. Evol. 37, 475–487. 642 (doi:10.1093/molbev/msz234) 643

644

645

646

(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 24, 2020. ; https://doi.org/10.1101/2020.07.22.208553doi: bioRxiv preprint

24

Competing interests statement 647

The authors declare no competing interests. 648

649

Authors’ contributions statement 650

DF conceived the study. MD conducted the molecular work and analysed the molecular data 651

with DF and input from JKC. NJA and CLO analysed the environmental data. DF and NJA 652

conducted the field work, interpreted the data and wrote the manuscript with input from all 653

authors. All authors gave final approval for publication and agree to be held accountable for 654

the work performed therein. 655

(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 24, 2020. ; https://doi.org/10.1101/2020.07.22.208553doi: bioRxiv preprint