Philippa Kaur

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.This is a summary of: Sasaki, M. et al. Greater evolutionary divergence of thermal limits within marine than terrestrial species. Nat. Clim. Change https://doi.org/10.1038/s41558-022-01534-y (2022). More

Honey bee colony loss and parasites across space and timeHoney bee colony loss strongly depends on spatio-temporal factors33,42, which in turn have to be jointly modeled with other stressors. Focusing on CONUS climatic regions, defined by the National Centers for Environmental Information40 (see Fig. 1), this is supported by the box plots in Fig. 2 which depict appropriately normalized honey bee colony loss (upper panel) and presence of V. destructor (lower panel) quarterly between 2015 and 2021. Specifically, Fig. 2a highlights that the first quarter generally accounts for a higher and more variable proportion of losses. Average losses are typically lower and less dispersed during the second quarter, and then tend to increase again during the third and fourth quarters. The Central region, which reports the highest median losses during the first quarter (larger than 20%) exemplifies this pattern, which is in line with existing studies that link overwintering with honey bee colony loss6,29,30,31,32,33,43. On the other hand, the West North Central region follows a different pattern, where losses are typically lower during the first quarter and peak during the third. This holds, albeit less markedly, also for Northwest and Southwest regions. These differing patterns are also depicted in Fig. 3, which shows the time series of normalized colony loss for each state belonging to Central and West North Central regions – with the smoothed conditional means highlighted in black and red, respectively. Figure 2b shows that also the presence of V. destructor tends to follow a specific pattern; in most regions it increases from the first to the third quarter, and then it decreases in the fourth – with the exception of the Southwest region, where it keeps increasing. This is most likely because most beekeepers try to get V. destructor levels low by fall, so that colonies are as healthy as possible going into winter, and also because of the population dynamics of V. destructor alongside honey bee colonies – i.e., their presence typically increases as the colony grows and has more brood cycles, since this parasite develops inside honey bee brood cells44,45. The West region (which encompasses only California since Nevada was missing in the honey bee dataset; see Data) reports high levels of V. destructor throughout the year, with very small variability. A comparison of Fig. 2a and b shows that honey bee colony loss and the presence of V. destructor tend to be higher than the corresponding medians during the third quarter, suggesting a positive association. This is further confirmed in Fig. 4, which shows a scatter plot of normalized colony loss against V. destructor presence, documenting a positive association in all quarters. Although with the data at hand we are not able to capture honey bee movement across states, as well as intra-quarter losses and honey production, these preliminary findings can be useful to support commercial beekeeper strategies and require further investigation.Figure 2Empirical distribution of honey bee (Apis mellifera) colony loss (a) and Varroa destructor presence (b) across quarters (the first one being January-March) and climatic regions; red dashed lines indicate the overall medians. (a) Box plots of normalized colony loss (number of lost colonies over the maximum number of colonies) for each quarter of 2015–2021 and each climatic region. At the contiguous United States level, this follows a stable pattern across the years, with higher and more variable losses during the first quarter (see Supplementary Figs. S2-S6), but some regions do depart from this pattern (e.g., West North Central). (b) Box plots of normalized V. destructor presence (number of colonies affected by V. destructor over the maximum number of colonies) for each quarter of 2015–2021 and each climatic region. The maximum number of colonies is defined as the number of colonies at the beginning of a quarter, plus all colonies moved into that region during the same quarter.Full size imageFigure 3Comparison of normalized honey bee (Apis mellifera) colony loss (number of lost colonies over the maximum number of colonies) between Central and West North Central climatic regions for each quarter of 2015–2021 (the first quarter being January-March). (a) Trajectory of each state belonging to Central (yellow) and West North Central (blue) climatic regions. (b) Smoothed conditional means for each of the two sets of curves based on a locally weighted running line smoother where the width of the sliding window is equal to 0.2 and corresponding standard error bands are based on a 0.95 confidence level46.Full size imageFigure 4Scatter plot of normalized honey bee (Apis mellifera) colony loss (number of lost colonies over the maximum number of colonies) against normalized Varroa destructor presence (number of colonies affected by V. destructor over the maximum number of colonies) for each state and each quarter of 2015–2021 (the first quarter being January-March). Points are color-coded by quarter, and ordinary least squares fits (with corresponding standard error bands based on a 0.95 confidence level) computed by quarter are superimposed to visualize the positive association.Full size imageUp-scaling weather dataThe data sets available to us for weather related variables had a much finer spatio-temporal resolution (daily and on a (4 times 4) kilometer grid) than the colony loss data (quarterly and at the state level). Therefore, we aggregated the former to match the latter. For similar data up-scaling tasks, sums or means are commonly employed to summarize the variables available at finer resolution47. The problem with aggregating data in such a manner is that one only preserves information on the “center” of the distributions – thus losing a potentially considerable amount of information. To retain richer weather related information in our study, we considered additional summaries capturing more complex characteristics, e.g., the tails of the distributions or their entropy, to ascertain whether they may help in predicting honey bee colony loss. Within each state and quarter we therefore computed, in addition to means, indexes such as standard deviation, skewness, kurtosis, (L_2)-norm (or energy), entropy and tail indexes48. This was done for minimum and maximum temperatures, as well as precipitation data (see Data processing for details).Next, as a first way to validate the proposed weather data up-scaling approach, we performed a likelihood ratio test between nested models. Specifically, we considered a linear regression for colony loss (see Statistical model) and compared an ordinary least squares fit comprising all the computed indexes as predictors (the full model) against one comprising only means and standard deviations (the reduced model). The test showed that the use of additional indexes provides a statistically significant improvement in the fit (p-(text {value}=0.03)). This test, which can be replicated for other choices of models and estimation methods (see Supplementary Table S5), supports the use of our up-scaling approach.Figure 5 provides a spatial representation of (normalized) honey bee colony losses and of three indexes relative to the minimum temperature distribution; namely, mean, kurtosis and skewness (these all turn out to be relevant predictors based on subsequent analyses; see Table 1). For each of the four quantities, the maps are color-coded by state based on the median of first quarter values over the period 2015-2021 (first quarters typically have the highest losses, but similar patterns can be observed for other quarters; see Supplementary Figs. S12-S14). Notably, the indexes capture characteristics of the within-state distributions of minimum temperatures that do vary geographically. For example, considering minimum temperature, skewness is an index that (broadly speaking) provides information on whether the data tends to accumulate at one end or the other of the observed range of minimum temperatures (i.e., a positive/negative skewness indicates that the data accumulates towards the lower/upper range, respectively). On the other hand, kurtosis is an index that captures the presence of “extreme” values in the tails of the data (i.e., a low/high value of kurtosis indicates that the tail minimum temperatures are relatively close/very far from the typical minimum temperatures). With this in mind, going back to Fig. 5, we can see that minimum temperatures in states in the north-west present large kurtosis (a prevalence of extreme values in the tails) and negative skewness (a tendency to accumulate towards the upper values of the minimum temperature range), while the opposite is true for states in the south-east. More generally, the mean minimum temperature separates northern vs southern states, kurtosis is higher for states located in the central band of the CONUS, and skewness separates western vs eastern states.We further note that the states with lower losses during the first quarter (e.g., Montana and Wyoming) do not report extreme values in any of the considered indexes. Although these states are generally characterized by low minimum temperatures, these are somewhat “stable” (they do not show marked kurtosis or skewness in their distributions) – perhaps allowing honey bees and beekeepers to adapt to more predictable conditions. On the other hand, states with higher losses during the first quarter such as New Mexico have higher minimum temperatures as well as marked kurtosis, and thus higher chances of extreme minimum temperatures – which may indeed affect honey bee behavior and colony loss. Overall, across all quarters of the years 2015-2021, we found that normalized colony losses and mean minimum temperatures are negatively associated (the Pearson correlation is -0.17 with a p-(text {value} More

Rearing O. biroi pupae in social isolation and collecting pupal fluidIn O. biroi colonies, larvae and pupae develop in discrete and synchronized cohorts26. Ten days after the first larvae had entered pupation in a large stock colony, the entire colony was anaesthetized using a CO2 pad, and white pupae were separated using a paintbrush. Pupae were individually placed in 0.2 ml PCR tubes with open lid. These tubes were then placed inside 1.5 ml Eppendorf tubes with 5 µl sterile water at the bottom to provide 100% relative humidity. The outer tubes were closed and kept in a climate room at 25 °C. The inner tube in this design prevents the pupa from drowning in the water reservoir. The outer tubes were kept closed throughout the experiment, except for once a day when the tubes were opened to remove pupal social fluid. Pulled glass capillaries were prepared as described elsewhere29, and used to remove and/or collect secretion droplets. We were careful to leave no remains of the secretion behind on the pupae or the inside of the tubes. To ensure that all secretion had been removed, pupae were taken out of the tube after fluid collection and briefly placed on a tissue paper to absorb any excess liquid. The inner tubes were replaced if needed—for example, if fluid traces were visible on the old tube after collection. Each pupa was checked daily for secretion (absent or present), onset of melanization and eclosion, and whether the pupa was alive (responding to touch). Control groups of 30 pupae and 30 adult ants from the same stock colony and cohort as the isolated pupae were placed in Petri dishes with a plaster of Paris floor, and the same parameters as for the isolated pupae were scored daily. Experiments ended when all pupae had either eclosed or died. Newly eclosed (callow) workers moved freely inside the tube and showed no abnormalities when put in a colony. A pupa was declared dead if it did not shed its pupal skin and did not respond to touch three days after all pupae in the control group had eclosed.To calculate the average secretion volume per secreting pupa (Fig. 1d), the total volume collected daily from a group of isolated pupae (142–166 pupae) was divided by the number of pupae from which fluid had been collected that day. The total volume was determined by multiplying the height of the fluid’s meniscus in the capillary by πr², where r is the inner radius of the capillary (0.29 mm). While pupae were secreting, pupal whole-body wash samples were collected daily. The pupae were removed from colonies with adults and washed promptly with 1500 µl LC–MS grade water. Whole-body wash samples were lyophilized and reconstituted in 15 µl LC–MS grade water.Collecting additional ant species and honeybees, rearing pupae in social isolation, and collecting pupal fluidsColonies of the ants N. flavipes, T. sessile, P. pennsylvanica and Lasius neoniger were collected in NY state, USA (Central Park, Manhattan; Pelham Bay Park, Bronx; Prospect Park, Brooklyn; and Woodstock). Solenopsis invicta colonies were collected in Athens, GA, USA. M. mexicanus colonies were collected in Piñon Hills, CA, USA. Colonies comprised of queens, workers and brood were maintained in the laboratory in airtight acrylic boxes with plaster of Paris floors. Colonies were fed a diet of insects (flies, crickets and mealworms). White pupae were socially isolated, cocoons were removed in the case of P. pennsylvanica, and secretion droplets were collected from melanized pupae as described for O. biroi. A. mellifera pupae of unknown age were socially isolated from hive fragments (A&Z Apiaries, USA) and reared as described for O biroi, except that the rearing temperature was set to 32 °C. Relative humidity was set to either 100% to replicate conditions used for the different ant species, or to 75% as recommended in the literature30.Injecting dye and tracking pupal fluidInjection needles were prepared as in previous studies31. Injections were performed using an Eppendorf Femtojet with a Narishige micromanipulator. The Femtojet was set to Pi 1000 hPa and Pc 60 hPa. Needles were broken by gently touching the capillary tip to the side of a glass slide. To inject, melanized pupae were placed on ‘Sticky note’ tape (Post-it), with the abdomen tip forward and the ventral side upward. Pupae were injected with blue food colouring (McCormick) into the exuvium for 1–2 s by gently piercing the pupal case at the abdominal tip with the needle. During successful injections, no fluid was discharged from the pupa when the needle was removed, and the moulting fluid inside the exuvium was immediately stained. Pupae were washed in water three times to remove any excess dye. Following injections, 10 pupae were reared in social isolation to confirm the secretion of dyed droplets. For experiments, injected pupae were transferred to colonies with adult ants (Figs. 1f and  4c) or to colonies with adult ants and larvae (Figs. 3b and  4c) to track the distribution of the pupal social fluid.After spending 24 h with dye-injected pupae, adults were taken out of the colony, briefly immersed in 95% ethanol, and transferred to PBS. Digestive systems were dissected in cold PBS and mounted in DAKO mounting medium. Crop and stomach images (Fig. 1f, inset and Fig. 4c, inset) were acquired with a Revolve microscope (Echo). Larvae are translucent, and the presence of dye in the digestive system can be assayed without dissection. Whole-body images of larvae were acquired with a Leica Z16 APO microscope equipped with a Leica DFC450 camera and Leica Application Suite version 4.12.0 (Leica Microsystems). In the experiment on larval growth (Fig. 3c), larval length was measured from images using ImageJ32.Occluding pupaeTen pupae were placed on double-sided tape on a glass coverslip with the ventral side up. The area between the pupae was covered with laser-cut filter paper to prevent adults from sticking to the tape. The pupae were then placed in a 5 cm diameter Petri dish with a moist plaster of Paris floor. To block pupal secretion, the tip of the gaster was occluded with a drop of oil-paint (Uni Paint Markers PX-20), which has no discernible toxic effect7. Secreting pupae received a drop of the same paint on their head to control for putative differences resulting from the paint. Pupae were left in isolation for one day before adults were added to the assay chamber.Behavioural tracking of adult preference assayVideos were recorded using BFS-U3-50S5C-C: 5.0 MP, 35 FPS, Sony IMX264, Colour cameras (FLIR) and the Motif Video Recording System (Loopbio). To assess adult preference (Fig. 1g), physical contact of adults with pupae was manually annotated for the first 10 min after the first adult had encountered (physically contacted) a pupa.Protein profilingWe extracted 30 µl of pupal social fluid and whole-body wash samples with 75:25:0.2 acetonitrile: methanol: formic acid. Extracts were vortexed for 10 min, centrifuged at 16,000g and 4 °C for 10 min, dried in a SpeedVac, and stored at −80 °C until they were analysed by LC–MS/MS.Protein pellets were dissolved in 8 M urea, 50 mM ammonium bicarbonate, and 10 mM dithiothreitol, and disulfide bonds were reduced for 1 h at room temperature. Alkylation was performed by adding iodoacetamide to a final concentration of 20 mM and incubating for 1 h at room temperature in the dark. Samples were diluted using 50 mM ammonium bicarbonate until the concentration of urea had reached 3.5 M, and proteins were digested with endopeptidase LysC overnight at room temperature. Samples were further diluted to bring the urea concentration to 1.5 M before sequencing-grade modified trypsin was added. Digestion proceeded for 6 h at room temperature before being halted by acidification with TFA and samples were purified using in-house constructed C18 micropurification tips.LC–MS/MS analysis was performed using a Dionex3000 nanoflow HPLC and a Q-Exactive HF mass spectrometer (both Thermo Scientific). Solvent A was 0.1% formic acid in water and solvent B was 80% acetonitrile, 0.1% formic acid in water. Peptides were separated on a 90-minute linear gradient at 300 nl min−1 across a 75 µm × 100 mm fused-silica column packed with 3 µm Reprosil C18 material (Dr. Maisch). The mass spectrometer operated in positive ion Top20 DDA mode at resolution 60 k/30 k (MS1/MS2) and AGC targets were 3 × 106/2 × 105 (MS1/MS2).Raw files were searched through Proteome Discoverer v.1.4 (Thermo Scientific) and spectra were queried against the O. biroi proteome using MASCOT with a 1% FDR applied. Oxidation of M and acetylation of protein N termini were applied as a variable modification and carbamidomethylation of C was applied as a static modification. The average area of the three most abundant peptides for a matched protein33 was used to gauge protein amounts within and between samples.Functional annotation and gene ontology enrichmentTo supplement the current functional annotation of the O. biroi genome34, the full proteome for canonical transcripts was retrieved from UniProtKB (UniProt release 2020_04) in FASTA format. We then applied the EggNog-Mapper tool35,36 (http://eggnog-mapper.embl.de, emapper version 1.0.3-35-g63c274b, EggNogDB version 2) using standard parameters (m diamond -d none –tax_scope auto –go_evidence non-electronic –target_orthologs all –seed_ortholog_evalue 0.001 –seed_ortholog_score 60 –query-cover 20 –subject-cover 0) to produce an expanded annotation for all GO trees (Molecular Function, Biological Process, Cellular Components). The list of proteins identified in the pupal fluid was evaluated for functional enrichment in these GO terms, P-values were adjusted with an FDR cut-off of 0.05, and the network plots were visualized using the clusterProfiler package37.Metabolite profilingFor bulk polar metabolite profiling, we used 10 µl aliquots of pupal social fluid and whole-body wash (pooled samples). For the time-series metabolite profiling, 1 µl of pupal social fluid and whole-body wash was used. Samples were extracted in 180 µl cold LC–MS grade methanol containing 1 μM of uniformly labelled 15N- and 13C-amino acid internal standards (MSK-A2-1.2, Cambridge Isotope Laboratories) and consecutive addition of 390 µl LC–MS grade chloroform followed by 120 µl of LC–MS grade water.The samples were vortexed vigorously for 10 min followed by centrifugation (10 min at 16,000g and 4 °C). The upper polar metabolite-containing layer was collected, flash frozen and SpeedVac-dried. Dried extracts were stored at −80 °C until LC–MS analysis.LC–MS was conducted on a Q-Exactive benchtop Orbitrap mass spectrometer equipped with an Ion Max source and a HESI II probe, which was coupled to a Vanquish UPLC system (Thermo Fisher Scientific). External mass calibration was performed using the standard calibration mixture every three days.Dried polar samples were resuspended in 60 µl 50% acetonitrile, and 5 µl were injected into a ZIC-pHILIC 150 × 2.1 mm (5 µm particle size) column (EMD Millipore). Chromatographic separation was achieved using the following conditions: buffer A was 20 mM ammonium carbonate, 0.1% (v/v) ammonium hydroxide (adjusted to pH 9.3); buffer B was acetonitrile. The column oven and autosampler tray were held at 40 °C and 4 °C, respectively. The chromatographic gradient was run at a flow rate of 0.150 ml min−1 as follows: 0–22 min: linear gradient from 90% to 40% B; 22–24 min: held at 40% B; 24–24.1 min: returned to 90% B; 24.1 −30 min: held at 90% B. The mass spectrometer was operated in full-scan, polarity switching mode with the spray voltage set to 3.0 kV, the heated capillary held at 275 °C, and the HESI probe held at 250 °C. The sheath gas flow was set to 40 units, the auxiliary gas flow was set to 15 units. The MS data acquisition was performed in a range of 55–825 m/z, with the resolution set at 70,000, the AGC target at 10 × 106, and the maximum injection time at 80 ms. Relative quantification of metabolite abundances was performed using Skyline Daily v 20.1 (MacCoss Lab) with a 2 ppm mass tolerance and a pooled library of metabolite standards to confirm metabolite identity (via data-dependent acquisition). Metabolite levels were normalized by the mean signal of 8 heavy 13C,15N-labelled amino acid internal standards (technical normalization).The raw data were searched for a targeted list of ~230 polar metabolites and the corresponding peaks were integrated manually using Skyline Daily software. We were able to assign peaks to 107 compounds based on high mass accuracy ( More

Hawley, W. A. The biology of Aedes albopictus. J. Am. Mosq. Control Assoc. 1, 1–39 (1988).CAS 

Google Scholar 
Benedict, M. Q., Levine, R. S., Hawley, W. A. & Lounibos, L. P. Spread of the tiger: global risk of invasion by the mosquito Aedes albopictus. Vector-Borne Zoonotic Dis. 7, 76–85 (2007).Article 
PubMed 

Google Scholar 
Paupy, C., Delatte, H., Bagny, L., Corbel, V. & Fontenille, D. Aedes albopictus, an arbovirus vector: From the darkness to the light. Microb. Infect. 11, 1177–1185 (2009).Article 
CAS 

Google Scholar 
Delatte, H. et al. Blood-feeding behavior of Aedes albopictus, a vector of Chikungunya on La Réunion. Vector-Borne Zoonotic Dis. 10, 249–258 (2010).Article 
PubMed 

Google Scholar 
Pereira-dos-Santos, T., Roiz, D., Lourenço-de-Oliveira, R. & Paupy, C. A systematic review: Is Aedes albopictus an efficient bridge vector for zoonotic arboviruses? Pathogens 9, 266 (2020).Article 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Gratz, N. Critical review of the vector status of Aedes albopictus. Med. Vet. Entomol. 18, 215–227 (2004).Article 
CAS 
PubMed 

Google Scholar 
Grard, G. et al. Zika virus in Gabon (Central Africa)—2007: A new threat from Aedes albopictus? PLoS Negl. Trop. Dis. 8, e2681 (2014).Article 
PubMed 
PubMed Central 

Google Scholar 
Lambrechts, L., Scott, T. W. & Gubler, D. J. Consequences of the expanding global distribution of Aedes albopictus for dengue virus transmission. PLoS Negl. Trop. Dis. 4, e646 (2010).Article 
PubMed 
PubMed Central 

Google Scholar 
Lounibos, L. P. & Kramer, L. D. Invasiveness of Aedes aegypti and Aedes albopictus and vectorial capacity for chikungunya virus. J. Infect. Dis. 214, S453–S458 (2016).Article 
PubMed 
PubMed Central 

Google Scholar 
European Centre for Disease Prevention and Control (ECDC). Vector Control with a Focus on Aedes aegypti and Aedes albopictus Mosquitoes: Literature Review and Analysis of Information (ECDC, Stockholm, Sweden, 2017).Tatem, A. J., Hay, S. I. & Rogers, D. J. Global traffic and disease vector dispersal. PNAS 103, 6242–6247 (2006).Article 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Lowe, S., Browne, M., Boudjelas, S. & De Poorter, M. 100 of the World’s Worst Invasive Alien Species: A Selection From the Global Invasive Species Database, Vol. 12 (Invasive Species Specialist Group, 2000).Diagne, C. et al. High and rising economic costs of biological invasions worldwide. Nature 592, 571–576 (2021).Article 
CAS 
PubMed 

Google Scholar 
Hulme, P. E. Trade, transport and trouble: Managing invasive species pathways in an era of globalization. J. Appl. Ecol. 46, 10–18 (2009).Article 

Google Scholar 
Marini, F., Caputo, B., Pombi, M., Tarsitani, G. & Della-Torre, A. Study of Aedes albopictus dispersal in Rome, Italy, using sticky traps in mark–release–recapture experiments. Med. Vet. Entomol. 24, 361–368 (2010).Article 
CAS 
PubMed 

Google Scholar 
Bonizzoni, M., Gasperi, G., Chen, X. & James, A. A. The invasive mosquito species Aedes albopictus: current knowledge and future perspectives. Trends Parasitol. 29, 460–468 (2013).Article 
PubMed 
PubMed Central 

Google Scholar 
Collantes, F. et al. Review of ten-years presence of Aedes albopictus in Spain 2004–2014: Known distribution and public health concerns. Parasit Vectors 8, 1–11 (2015).Article 

Google Scholar 
Aranda, C., Eritja, R. & Roiz, D. First record and establishment of the mosquito Aedes albopictus in Spain. Med. Vet. Entomol. 20, 150–152 (2006).Article 
CAS 
PubMed 

Google Scholar 
Giménez, N. et al. Introduction of Aedes albopictus in Spain: A new challenge for public health. Gac. Sanit. 21, 25–28 (2007).Article 
PubMed 

Google Scholar 
European Centre for Disease Prevention and Control and European Food Safety Authority. Mosquito maps [internet]. Stockholm: ECDC. https://ecdc.europa.eu/en/disease-vectors/surveillance-and-disease-data/mosquito-maps (2022).Shigesada, N. & Kawasaki, K. Biological Invasions: Theory and Practice (Oxford University Press, 1997).
Google Scholar 
Puth, L. M. & Post, D. M. Studying invasion: Have we missed the boat? Ecol. Lett. 8, 715–721 (2005).Article 

Google Scholar 
Leung, B. et al. An ounce of prevention or a pound of cure: Bioeconomic risk analysis of invasive species. Proc. R Soc. Lond. Ser. B Biol. Sci. 269, 2407–2413 (2002).Article 

Google Scholar 
Lounibos, L. P. Invasions by insect vectors of human disease. Annu. Rev. Entomol. 47, 233–266 (2002).Article 
CAS 
PubMed 

Google Scholar 
Manni, M. et al. Genetic evidence for a worldwide chaotic dispersion pattern of the arbovirus vector, Aedes albopictus. PLoS Negl. Trop. Dis. 11, e0005332 (2017).Article 
PubMed 
PubMed Central 

Google Scholar 
Roiz, D. et al. Integrated Aedes management for the control of Aedes-borne diseases. PLoS Negl. Trop. Dis. 12, e0006845 (2018).Article 
PubMed 
PubMed Central 

Google Scholar 
Lühken, R. et al. Microsatellite typing of Aedes albopictus (Diptera: Culicidae) populations from Germany suggests regular introductions. Infect. Genet. Evol. 81, 104237 (2020).Article 
PubMed 

Google Scholar 
Battaglia, V. et al. The worldwide spread of the tiger mosquito as revealed by mitogenome haplogroup diversity. Front. Genet. 7, 208 (2016).Article 
PubMed 
PubMed Central 

Google Scholar 
Medley, K. A., Jenkins, D. G. & Hoffman, E. A. Human-aided and natural dispersal drive gene flow across the range of an invasive mosquito. Mol. Ecol. 24, 284–295 (2015).Article 
PubMed 

Google Scholar 
Eritja, R., Palmer, J. R., Roiz, D., Sanpera-Calbet, I. & Bartumeus, F. Direct evidence of adult Aedes albopictus dispersal by car. Sci. Rep. 7, 1–15 (2017).Article 
CAS 

Google Scholar 
Sherpa, S. et al. Unravelling the invasion history of the Asian tiger mosquito in Europe. Mol. Ecol. 28, 2360–2377 (2019).Article 
PubMed 

Google Scholar 
Swan, T. et al. A literature review of dispersal pathways of Aedes albopictus across different spatial scales: Implications for vector surveillance. Parasit Vectors 15, 1–13 (2022).Article 

Google Scholar 
Ballard, J. W. O. & Whitlock, M. C. The incomplete natural history of mitochondria. Mol. Ecol. 13, 729–744. https://doi.org/10.1046/j.1365-294X.2003.02063.x (2004).Article 
PubMed 

Google Scholar 
Toews, D. P. L. & Brelsford, A. The biogeography of mitochondrial and nuclear discordance in animals. Mol. Ecol. 21, 3907–3930. https://doi.org/10.1111/j.1365-294X.2012.05664.x (2012).Article 
CAS 
PubMed 

Google Scholar 
Hurst, G. D. & Jiggins, F. M. Problems with mitochondrial DNA as a marker in population, phylogeographic and phylogenetic studies: The effects of inherited symbionts. Proc. R. Soc. B: Biol. Sci. 272, 1525–1534 (2005).Article 
CAS 

Google Scholar 
Cariou, M., Duret, L. & Charlat, S. The global impact of Wolbachia on mitochondrial diversity and evolution. J. Evol. Biol. 30, 2204–2210 (2017).Article 
CAS 
PubMed 

Google Scholar 
Zug, R. & Hammerstein, P. Still a host of hosts for Wolbachia: analysis of recent data suggests that 40% of terrestrial arthropod species are infected. PLoS ONE 7, e38544 (2012).Article 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Weinert, L. A., Araujo-Jnr, E. V., Ahmed, M. Z. & Welch, J. J. The incidence of bacterial endosymbionts in terrestrial arthropods. Proc. R. Soc. B: Biol. Sci. 282, 20150249 (2015).Article 

Google Scholar 
Goubert, C., Minard, G., Vieira, C. & Boulesteix, M. Population genetics of the Asian tiger mosquito Aedes albopictus, an invasive vector of human diseases. Heredity 117, 125–134 (2016).Article 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Western, D. Human-modified ecosystems and future evolution. PNAS 98, 5458–5465 (2001).Article 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Pech-May, A. et al. Population genetics and ecological niche of invasive Aedes albopictus in Mexico. Acta Trop. 157, 30–41 (2016).Article 
PubMed 

Google Scholar 
Vargo, E. L. et al. Hierarchical genetic analysis of German cockroach (Blattella germanica) populations from within buildings to across continents. PLoS ONE 9, e102321 (2014).Article 
PubMed 
PubMed Central 

Google Scholar 
von Beeren, C., Stoeckle, M. Y., Xia, J., Burke, G. & Kronauer, D. J. Interbreeding among deeply divergent mitochondrial lineages in the American cockroach (Periplaneta americana). Sci. Rep. 5, 1–7 (2015).
Google Scholar 
Tseng, S.-P. et al. Genetic diversity and Wolbachia infection patterns in a globally distributed invasive ant. Front. Genet. 10, 838 (2019).Article 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Wesson, D. M., Porter, C. H. & Collins, F. H. Sequence and secondary structure comparisons of ITS rDNA in mosquitoes (Diptera: Culicidae). Mol. Phylogen. Evol. 1, 253–269 (1992).Article 
CAS 

Google Scholar 
Mishra, S., Sharma, G., Das, M. K., Pande, V. & Singh, O. P. Intragenomic sequence variations in the second internal transcribed spacer (ITS2) ribosomal DNA of the malaria vector Anopheles stephensi. PLoS ONE 16, e0253173 (2021).Article 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Artigas, P. et al. Aedes albopictus diversity and relationships in south-western Europe and Brazil by rDNA/mtDNA and phenotypic analyses: ITS-2, a useful marker for spread studies. Parasit Vectors 14, 1–23 (2021).Article 

Google Scholar 
Armbruster, P. et al. Infection of New-and Old-World Aedes albopictus (Diptera: Culicidae) by the intracellular parasite Wolbachia: implications for host mitochondrial DNA evolution. J. Med. Entomol. 40, 356–360 (2003).Article 
PubMed 

Google Scholar 
Maia, R., Scarpassa, V. M., Maciel-Litaiff, L. & Tadei, W. P. Reduced levels of genetic variation in Aedes albopictus (Diptera: Culicidae) from Manaus, Amazonas State, Brazil, based on analysis of the mitochondrial DNA ND5 gene. Gen. Mol. Res. 2000, 998–1007 (2009).Article 

Google Scholar 
Birungi, J. & Munstermann, L. E. Genetic structure of Aedes albopictus (Diptera: Culicidae) populations based on mitochondrial ND5 sequences: Evidence for an independent invasion into Brazil and United States. Ann. Entomol. Soc. Am. 95, 125–132 (2002).Article 
CAS 

Google Scholar 
Kambhampati, S. & Rai, K. S. Mitochondrial DNA variation within and among populations of the mosquito Aedes albopictus. Genome 34, 288–292 (1991).Article 
CAS 
PubMed 

Google Scholar 
Werren, J. H., Baldo, L. & Clark, M. E. Wolbachia: master manipulators of invertebrate biology. Nat. Rev. Microbiol. 6, 741–751 (2008).Article 
CAS 
PubMed 

Google Scholar 
Wiwatanaratanabutr, I. Geographic distribution of wolbachial infections in mosquitoes from Thailand. J. Invertebr. Pathol. 114, 337–340 (2013).Article 
PubMed 

Google Scholar 
Carvajal, T. M., Hashimoto, K., Harnandika, R. K., Amalin, D. M. & Watanabe, K. Detection of Wolbachia in field-collected Aedes aegypti mosquitoes in metropolitan Manila, Philippines. Parasit. Vectors 12, 1–9 (2019).Article 

Google Scholar 
Atyame, C. M., Delsuc, F., Pasteur, N., Weill, M. & Duron, O. Diversification of Wolbachia endosymbiont in the Culex pipiens mosquito. Mol. Biol. Evol. 28, 2761–2772 (2011).Article 
CAS 
PubMed 

Google Scholar 
Damiani, C. et al. Wolbachia in Aedes koreicus: Rare detections and possible implications. Insects 13, 216 (2022).Article 
PubMed 
PubMed Central 

Google Scholar 
Jiggins, F. M. Male-killing Wolbachia and mitochondrial DNA: Selective sweeps, hybrid introgression and parasite population dynamics. Genetics 164, 5–12 (2003).Article 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Schuler, H. et al. The hitchhiker’s guide to Europe: The infection dynamics of an ongoing Wolbachia invasion and mitochondrial selective sweep in Rhagoletis cerasi. Mol. Ecol. 25, 1595–1609 (2016).Article 
PubMed 
PubMed Central 

Google Scholar 
Ross, P. A., Ritchie, S. A., Axford, J. K. & Hoffmann, A. A. Loss of cytoplasmic incompatibility in Wolbachia-infected Aedes aegypti under field conditions. PLoS Negl. Trop. Dis. 13, e0007357 (2019).Article 
PubMed 
PubMed Central 

Google Scholar 
Avise, J. C. Phylogeography: The history and formation of species (Harvard University Press, 2000).Book 

Google Scholar 
Rokas, A., Atkinson, R. J., Brown, G. S., West, S. A. & Stone, G. N. Understanding patterns of genetic diversity in the oak gallwasp Biorhiza pallida: Demographic history or a Wolbachia selective sweep? Heredity 87, 294–304 (2001).Article 
CAS 
PubMed 

Google Scholar 
Porretta, D., Mastrantonio, V., Bellini, R., Somboon, P. & Urbanelli, S. Glacial history of a modern invader: Phylogeography and species distribution modelling of the Asian tiger mosquito Aedes albopictus. PLoS ONE 7, e44515. https://doi.org/10.1371/journal.pone.0044515 (2012).Article 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Motoki, M. T. et al. Population genetics of Aedes albopictus (Diptera: Culicidae) in its native range in Lao People’s Democratic Republic. Parasit. Vectors 12, 1–12 (2019).Article 
CAS 

Google Scholar 
Zhong, D. et al. Genetic analysis of invasive Aedes albopictus populations in Los Angeles County, California and its potential public health impact. PLoS ONE 8, e68586 (2013).Article 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Usmani-Brown, S., Cohnstaedt, L. & Munstermann, L. E. Population genetics of Aedes albopictus (Diptera: Culicidae) invading populations, using mitochondrial nicotinamide adenine dinucleotide dehydrogenase subunit 5 sequences. Ann. Entomol. Soc. Am. 102, 144–150 (2009).Article 
CAS 
PubMed 

Google Scholar 
Mousson, L. et al. Phylogeography of Aedes (Stegomyia) aegypti (L.) and Aedes (Stegomyia) albopictus (Skuse) (Diptera: Culicidae) based on mitochondrial DNA variations. Genet. Res. 86, 1–11 (2005).Article 
CAS 
PubMed 

Google Scholar 
Bazin, E., Glémin, S. & Galtier, N. Population size does not influence mitochondrial genetic diversity in animals. Science 312, 570–572. https://doi.org/10.1126/science.1122033 (2006).Article 
CAS 
PubMed 

Google Scholar 
Dowling, D. K., Friberg, U. & Lindell, J. Evolutionary implications of non-neutral mitochondrial genetic variation. Ecol. Evol. 23, 546–554 (2008).Article 

Google Scholar 
Montero-Pau, J., Gómez, A. & Muñoz, J. Application of an inexpensive and high-throughput genomic DNA extraction method for the molecular ecology of zooplanktonic diapausing eggs. Limnol. Oceanogr. Methods 6, 218–222 (2008).Article 
CAS 

Google Scholar 
Porter, C. H. & Collins, F. H. Species-diagnostic differences in a ribosomal DNA internal transcribed spacer from the sibling species Anopheles freeborni and Anopheles hermsi (Diptera: Culicidae). Am. J. Trop. Med. 45, 271–279 (1991).Article 
CAS 

Google Scholar 
Folmer, O., Black, M., Hoeh, W., Lutz, R. & Vrijenhoek, R. DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Mol. Mar. Biol. Biotechnol. 3, 294–299 (1994).CAS 
PubMed 

Google Scholar 
Prosser, S., Martínez-Arce, A. & Elías-Gutiérrez, M. A new set of primers for COI amplification from freshwater microcrustaceans. Mol. Ecol. Resour. 13, 1151–1155 (2013).CAS 
PubMed 

Google Scholar 
Ivanova, N. V., Zemlak, T. S., Hanner, R. H. & Hebert, P. D. Universal primer cocktails for fish DNA barcoding. Mol. Ecol. Notes 7, 544–548 (2007).Article 
CAS 

Google Scholar 
Kumar, S., Stecher, G. & Tamura, K. MEGA7: Molecular evolutionary genetics analysis version 7.0 for bigger datasets. Mol. Biol. Evol. 33, 1870–1874 (2016).Article 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Zhou, W., Rousset, F. & O’Neill, S. Phylogeny and PCR–based classification of Wolbachia strains using wsp gene sequences. Proc. R Soc. Lond. Ser. B Biol. Sci. 265, 509–515 (1998).Article 
CAS 

Google Scholar 
Braig, H. R., Zhou, W., Dobson, S. L. & O’Neill, S. L. Cloning and characterization of a gene encoding the major surface protein of the bacterial endosymbiont Wolbachia pipientis. J. Bacteriol. 180, 2373–2378 (1998).Article 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Hu, Y. et al. Identification and molecular characterization of Wolbachia strains in natural populations of Aedes albopictus in China. Parasit. Vectors 13, 1–14 (2020).
Google Scholar 
Heddi, A., Grenier, A.-M., Khatchadourian, C., Charles, H. & Nardon, P. Four intracellular genomes direct weevil biology: Nuclear, mitochondrial, principal endosymbiont, and Wolbachia. PNAS 96, 6814–6819 (1999).Article 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Rozas, J. et al. DnaSP 6: DNA sequence polymorphism analysis of large data sets. Mol. Biol. Evol. 34, 3299–3302 (2017).Article 
CAS 
PubMed 

Google Scholar 
Salzburger, W., Ewing, G. B. & Von Haeseler, A. The performance of phylogenetic algorithms in estimating haplotype genealogies with migration. Mol. Ecol. 20, 1952–1963 (2011).Article 
PubMed 

Google Scholar 
Stamatakis, A. RAxML-VI-HPC: Maximum likelihood-based phylogenetic analyses with thousands of taxa and mixed models. Bioinformatics 22, 2688–2690 (2006).Article 
CAS 
PubMed 

Google Scholar 
Darriba, D., Taboada, G. L., Doallo, R. & Posada, D. jModelTest 2: More models, new heuristics and parallel computing. Nat. Methods 9, 772. https://doi.org/10.1038/nmeth.2109 (2012).Article 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Gower, J. C. Some distance properties of latent root and vector methods used in multivariate analysis. Biometrika 53, 325–338 (1966).Article 
MathSciNet 
MATH 

Google Scholar 
R Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing. http://www.R-project.org/. (2021).Hijmans, R. J., Williams, E., Vennes, C. & Hijmans, M. R. J. Package ‘geosphere’. Spher. Trigon. 1, 5 (2017).
Google Scholar 
Palmer, J. R. et al. Citizen science provides a reliable and scalable tool to track disease-carrying mosquitoes. Nat. Commun. 8, 1–13 (2017).Article 

Google Scholar 
Mantel, N. & Valand, R. S. A technique of nonparametric multivariate analysis. Biometrics 1970, 547–558 (1970).Article 

Google Scholar 
Goslee, S. C. & Urban, D. L. The ecodist package for dissimilarity-based analysis of ecological data. J. Stat. Softw. 22, 1–19 (2007).Article 

Google Scholar 
Stewart, C. Zero-inflated beta distribution for modeling the proportions in quantitative fatty acid signature analysis. J. Appl. Stat. 40, 985–992 (2013).Article 
MathSciNet 
MATH 

Google Scholar 
Figueroa-Zúñiga, J. I., Arellano-Valle, R. B. & Ferrari, S. L. Mixed beta regression: A Bayesian perspective. Comput. Stat. Data Anal. 61, 137–147 (2013).Article 
MathSciNet 
MATH 

Google Scholar 
Branscum, A. J., Johnson, W. O. & Thurmond, M. C. Bayesian beta regression: Applications to household expenditure data and genetic distance between foot-and-mouth disease viruses. Aust. N. Z. J. Stat. 49, 287–301 (2007).Article 
MathSciNet 
MATH 

Google Scholar 
Ospina, R. & Ferrari, S. L. Inflated beta distributions. Stat. Pap. 51, 111–126 (2010).Article 
MathSciNet 
MATH 

Google Scholar 
Chung, H. & Beretvas, S. N. The impact of ignoring multiple membership data structures in multilevel models. Br. J. Math. Stat. Psychol. 65, 185–200 (2012).Article 
MathSciNet 
PubMed 
MATH 

Google Scholar  More

Author notesMikayla M. ClarkPresent address: University of Tennessee, Knoxville, TN, USAMichael W. SneddonPresent address: Predicine, Inc., Hayward, CA, USARoman SutorminPresent address: Google, Inc., San Francisco, CA, USAAuthors and AffiliationsLawrence Berkeley National Laboratory, Berkeley, CA, USADylan Chivian, Sean P. Jungbluth, Paramvir S. Dehal, Elisha M. Wood-Charlson, Richard S. Canon, Gavin A. Price, William J. Riehl, Michael W. Sneddon, Roman Sutormin & Adam P. ArkinOak Ridge National Laboratory, Oak Ridge, TN, USABenjamin H. Allen, Mikayla M. Clark, Miriam L. Land & Robert W. CottinghamArgonne National Laboratory, Lemont, IL, USATianhao Gu, Qizhi Zhang & Chris S. HenryAuthorsDylan ChivianSean P. JungbluthParamvir S. DehalElisha M. Wood-CharlsonRichard S. CanonBenjamin H. AllenMikayla M. ClarkTianhao GuMiriam L. LandGavin A. PriceWilliam J. RiehlMichael W. SneddonRoman SutorminQizhi ZhangRobert W. CottinghamChris S. HenryAdam P. ArkinCorresponding authorsCorrespondence to
Dylan Chivian or Adam P. Arkin. More

arising from Y. Wang et al. Nature https://doi.org/10.1038/s41586-021-04016-x (2021)A unique challenge for environmental DNA (eDNA)-based palaeoecological reconstructions and extinction estimates is that organisms can contribute DNA to sediments long after their death. Recently, Wang et al.1 discovered mammoth eDNA in sediments that are between approximately 4.6 and 7 thousand years (kyr) younger than the most recent mammoth fossils in North America and Eurasia, which they interpreted as mammoths surviving on both continents into the Middle Holocene epoch. Here we present an alternative explanation for these offsets: the slow decomposition of mammoth tissues on cold Arctic landscapes is responsible for the release of DNA into sediments for thousands of years after mammoths went extinct. eDNA records are important palaeobiological archives, but the mixing of undatable DNA from long-dead organisms into younger sediments complicates the interpretation of eDNA, particularly from cold and high-latitude systems.All animal tissues, including faeces, contribute DNA to eDNA records2, but the durations across which tissues can contribute genetic information must vary depending on tissue type and local rates of destruction and decomposition. On high-latitude landscapes, soft tissues and skeletal remains of large mammals may persist, unburied, for millennia3,4,5. For example, unburied antlers of caribou (Rangifer tarandus) from Svalbard (Norway) and Ellesmere Island (Canada) have been dated3,4 to between 1 and 2 cal kyr bp (calibrated kyr before present). Elephant seal (Mirounga leonina) remains near the Antarctic coastline5,6 can persist for more than 5,000 years. This is in contrast to bones in warmer settings, which persist for only centuries or decades7,8. Because bones are particularly resistant to decay, quantifying how their persistence changes across environments enables us to constrain the durations that dead individuals generally contribute to eDNA archives. To do this, we consolidated data on the oldest radiocarbon-dated surface-collected bones from different ecosystems. We included bones that we are reasonably confident persisted without being completely buried (‘never buried’), and bones for which exhumation cannot be confidently excluded (‘potentially never buried’). Pairing bone persistence with mean annual temperatures (MAT) from their sample localities, we find a strong link between the local temperature and the logged duration of bone persistence (Fig. 1, never buried bones: R2 = 0.94, P  More

Department of Zoology, University of Cambridge, Cambridge, UKYucheng Wang, Bianca De Sanctis, Ruairidh Macleod, Daniel Money & Eske WillerslevLundbeck Foundation GeoGenetics Centre, Globe Institute, University of Copenhagen, Copenhagen, DenmarkYucheng Wang, Ana Prohaska, Jialu Cao, Antonio Fernandez-Guerra, James Haile, Kurt H. Kjær, Thorfinn Sand Korneliussen, Nicolaj Krog Larsen, Ruairidh Macleod, Hugh McColl, Mikkel Winther Pedersen, Fernando Racimo, Alexandra Rouillard, Anthony H. Ruter, Lasse Vinner, David J. Meltzer & Eske WillerslevALPHA, State Key Laboratory of Tibetan Plateau Earth System, Environment and Resources (TPESER), Institute of Tibetan Plateau Research (ITPCAS), Chinese Academy of Sciences (CAS), Beijing, ChinaYucheng WangKey Laboratory of Western China’s Environmental Systems (Ministry of Education), College of Earth and Environmental Science, Lanzhou University, Lanzhou, ChinaHaoran DongGénomique Métabolique, Genoscope, Institut François Jacob, CEA, CNRS, Univ Evry, Université Paris-Saclay, Evry, FranceAdriana Alberti, France Denoeud & Patrick WinckerInstitute for Integrative Biology of the Cell (I2BC), Université Paris-Saclay, CEA, CNRS, Gif-sur-Yvette, FranceAdriana AlbertiThe Arctic University Museum of Norway, UiT—The Arctic University of Norway, Tromsø, NorwayInger Greve Alsos, Eric Coissac, Galina Gusarova, Youri Lammers & Marie Kristine Føreid MerkelDepartment of Geography and Environment, University of Hawaii, Honolulu, HI, USADavid W. BeilmanDepartment of Geosciences and Natural Resource Management, University of Copenhagen, Copenhagen, DenmarkAnders A. BjørkInstitute of Earth Sciences, St Petersburg State University, St Petersburg, RussiaAnna A. Cherezova & Grigory B. FedorovArctic and Antarctic Research Institute, St Petersburg, RussiaAnna A. Cherezova & Grigory B. FedorovUniversité Grenoble-Alpes, Université Savoie Mont Blanc, CNRS, LECA, Grenoble, FranceEric CoissacDepartment of Genetics, University of Cambridge, Cambridge, UKBianca De Sanctis & Richard DurbinCarlsberg Research Laboratory, Copenhagen V, DenmarkChristoph Dockter & Birgitte SkadhaugeSchool of Geography and Environmental Science, University of Southampton, Southampton, UKMary E. EdwardsAlaska Quaternary Center, University of Alaska Fairbanks, Fairbanks, AK, USAMary E. EdwardsSchool of Environment, Earth and Ecosystem Sciences, The Open University, Milton Keynes, UKNeil R. Edwards & Philip B. HoldenCenter for the Environmental Management of Military Lands, Colorado State University, Fort Collins, CO, USAJulie EsdaleDepartment of Earth and Atmospheric Sciences, University of Alberta, Edmonton, Alberta, CanadaDuane G. FroeseFaculty of Biology, St Petersburg State University, St Petersburg, RussiaGalina GusarovaDepartment of Glaciology and Climate, Geological Survey of Denmark and Greenland, Copenhagen K, DenmarkKristian K. KjeldsenDepartment of Earth Science, University of Bergen, Bergen, NorwayJan Mangerud & John Inge SvendsenBjerknes Centre for Climate Research, Bergen, NorwayJan Mangerud & John Inge SvendsenDepartment of Geology, Quaternary Sciences, Lund University, Lund, SwedenPer MöllerCenter for Macroecology, Evolution and Climate, Globe Institute, University of Copenhagen, Copenhagen Ø, DenmarkDavid Nogués-Bravo, Hannah Lois Owens & Carsten RahbekCentre d’Anthropobiologie et de Génomique de Toulouse, Faculté de Médecine Purpane, Université Paul Sabatier, Toulouse, FranceLudovic OrlandoCenter for Global Mountain Biodiversity, Globe Institute, University of Copenhagen, Copenhagen, DenmarkHannah Lois Owens & Carsten RahbekGates of the Arctic National Park and Preserve, US National Park Service, Fairbanks, AK, USAJeffrey T. RasicDepartment of Geosciences, UiT—The Arctic University of Norway, Tromsø, NorwayAlexandra RouillardZoological Institute, Russian academy of sciences, St Petersburg, RussiaAlexei TikhonovResource and Environmental Research Center, Chinese Academy of Fishery Sciences, Beijing, ChinaYingchun XingCollege of Plant Science, Jilin University, Changchun, Jilin, ChinaYubin ZhangDepartment of Anthropology, Southern Methodist University, Dallas, TX, USADavid J. MeltzerWellcome Trust Sanger Institute, Wellcome Genome Campus, Cambridge, UKEske WillerslevMARUM, University of Bremen, Bremen, GermanyEske WillerslevAll authors contributed to the conception of the presented ideas. Y.W. and H.D. analysed the data. Y.W., D.J.M., A.P. and E.W. wrote the paper with inputs from all authors. More

Study siteThe Cariaco Basin, located on the continental shelf off Venezuela, is a large (about 160 km long and about 65 km wide) depression, composed of two approximately 1,400-m-deep sub-basins. It is partially isolated from the Caribbean Sea by a series of sills with depths of less than 150 m (ref. 47). This limits renewal of deep water in the basin and, paired with the high oxygen demand resulting from intense surface primary productivity, leads to anoxic waters below a depth of about 275 m at present47,48.The marked seasonality in the Cariaco Basin, combined with anoxic bottom waters that effectively prevent bioturbation, results in the accumulation of annually laminated (varved) sediments. As sediments are varved for the last deglaciation and the Holocene, and because of the sensitivity of the area to climate change, they are considered to be one of the most valuable high-resolution marine climate archives and have been successfully used to study climate variability in the tropics3,11,16,17,18. Varve thickness is about 1 mm or more during the YD–Holocene transition18.Core and age modelCore MD03-2621 was retrieved during IMAGES cruise XI (PICASSO) aboard R/V Marion Dufresne in 2003 (Laj and Shipboard Party 2004). Cariaco cores have been collected under the regulations of the Ocean Drilling Program and the IMAGES coring programme. In this study, data from depths between 480 and 540 cm below the seafloor are presented, encompassing the YD–Holocene transition. A detailed age model for core MD03-2621 was established by Deplazes et al.11 and is based on the cross-correlation of total reflectance to dated colour records from the Cariaco Basin49,50. For the studied interval, the original age model is based on a floating varve chronology anchored to tree ring data by matching 14C data49. The age model for core MD03-2621 was further fine-tuned by correlation of reflectance data to the NGRIP ice core δ18O record on the GICC05 age scale11. The transition from the YD to the Holocene is characterized by a decrease in the sedimentation rate from 1.4 to 0.5 mm year−1.To account for possible depth offsets during storage and subsampling, we matched sediment colour data expressed as greyscale (GS) to the reflectance data from Deplazes et al.11 with the software QAnalySeries51. To enable comparison with our record, ages in Lea et al.3 were corrected for the age difference between the sediment-colour-based midpoint of the YD–Holocene transition in their record (11.56 kyr b2k) and in data from Deplazes et al.11 (11.673 kyr b2k). The start and end of the change in reflectance were determined by the RAMPFIT software52.Sample preparationSamples for MSI of molecular proxies were prepared as described in Alfken et al.53: the original core was subsampled by LL channels, from which X-ray pictures (Hewlett-Packard Faxitron 43855A X-ray cabinet) and high-resolution digital images (smart-CIS 1600 Line Scanner) were obtained. The LL channels were then cut into 5-cm pieces, which were subsequently freeze-dried, embedded in a gelatin:carboxymethyl cellulose (4%:1%) mixture and thin-sectioned on a Microm HM 505 E cryomicrotome. From each piece, one 60-µm-thick and one 100-µm-thick, longitudinal slice (spanning the whole 5 cm piece) were prepared and affixed to indium-tin-oxide-coated glass slides (Bruker Daltonik, Bremen, Germany) for MSI and elemental mapping, respectively. Slices for MSI were further amended with a fullerite matrix54.For all slices, a high-resolution picture was taken on a M4 Tornado micro-X-ray fluorescence spectroscopy system (Bruker Nano Analytics). This picture was used as a reference to set up elemental mapping and MSI analysis, and also for the 2D comparison of elemental and proxy data to sediment colour. Sediment colour is expressed as GS value. To account for differences between single slices, ΔGS was calculated as the difference between a value and the median GS of each individual slice. Very low GS values corresponding to areas devoid of sediment, identified by a black background, were excluded from analysis.Elemental mappingElemental mapping of 100-µm-thick slices was performed on a M4 Tornado micro-X-ray fluorescence spectroscopy system (Bruker Nano Analytics) equipped with a micro-focused Rh source (50 kV, 600 µA) with a polycapillary optic. Measurements were conducted under vacuum, with a resolution of 50 µm, two scans per spot and a scan time of 5 ms per scan. Data were initially processed and visualized with M4 Tornado Software version 1.3. XY matrices of relevant elements and sediment colour were imported into MATLAB (R2016b) for further processing. To assess the correspondence between sediment colour and elemental composition, for each 5-cm piece, signal intensities of Ca, Fe, Ti and Si in single spots were binned according to ΔGS and average intensities were calculated for each bin (Extended Data Fig. 5). The bin size was 5 GS units.Molecular proxy analysis by MSIMSI was carried out on a 7T solariX XR Fourier transform ion cyclotron resonance mass spectrometer coupled to a matrix-assisted laser desorption/ionization source equipped with a Smartbeam II laser (Bruker Daltonik, Bremen, Germany). Analyses were performed in positive ionization mode selecting for a continuous accumulation of selective ions window of m/z 554 ± 12. Spectra were acquired with 25% data reduction to limit data size. Spatial resolution was obtained by rastering the ionizing laser across the sample in a defined rectangular area at a 100-µm spot distance. Considering laminae thickness in the millimetre range18, such raster resolution is suited for seasonally resolved SST reconstruction. Settings for laser power, frequency and number of shots were adjusted for optimal signal intensities before each measurement; typical values were 250 shots with 200 Hz frequency and 60% laser power. External mass calibration was performed in electrospray ionization mode with sodium trifluoroacetate (Sigma-Aldrich). Each spectrum was also calibrated after data acquisition by an internal lock mass calibration using the Na+ adduct of pyropheophorbide a (m/z 557.2523), a chlorophyll a derivative generally present in relatively young marine sediments. Around 20,000 individual spots were thereby obtained for every 5-cm slice, each spot containing information on the abundance of diunsaturated and triunsaturated C37 alkenones needed to calculate the ({{rm{U}}}_{37}^{{{rm{K}}}^{{prime} }}) SST proxy.We provide a two-pronged approach to decode SST proxy information: (1) a downcore ({{rm{U}}}_{37}^{{{rm{K}}}^{{prime} }}) profile is obtained by pooling alkenone data from coeval horizons, and results in SST reconstructions with annual resolution, and (2) 2D images of alkenone distribution are examined in conjunction with maps of sediment colour and elemental distribution to filter single-spot alkenone data for season of deposition.SST reconstruction with yearly resolutionFor the downcore profile, MSI data were referenced to the X-ray image by the identification of three teaching points per 5-cm piece. Afterwards, the X-ray image was corrected for tilting of laminae in the LL channels. This was done by identification of single laminae in the X-ray image and selection of a minimum of four tie points per lamina. A detailed description can be found in Alfken et al.9. After applying the corresponding age model, downcore profiles were established with 1-year resolution: the intensity of the two alkenone species relevant to the ({{rm{U}}}_{37}^{{{rm{K}}}^{{prime} }}) proxy (C37:2 and C37:3) were recorded for each individual laser spot and filtered for a signal-to-noise threshold of 3. Only spots in which both compounds were detected were further considered. Intensity values were then summed over the depth corresponding to 1 year. By pooling proxy data into 1-year horizons, the effect of changing sedimentation rate and, thereby, changing downcore resolution is minimized. If at least ten spots presenting both compounds were available for a single horizon, data quality criteria were satisfied54 and a ({{rm{U}}}_{37}^{{{rm{K}}}^{{prime} }}) value was calculated as defined by Prahl and Wakeham22:$${{rm{U}}}_{37}^{{{rm{K}}}^{{prime} }}=frac{{text{C}}_{37:2}}{{text{C}}_{37:2}+{text{C}}_{37:3}}$$
(1)
To apply the gas chromatography (GC)-based calibrations for the ({{rm{U}}}_{37}^{{{rm{K}}}^{{prime} }}) proxy, MSI-based data were converted to GC equivalents. Therefore, after MSI, sediment slices were extracted for conventional proxy analysis. Sediment was scraped off the slide and extracted following a modified Bligh and Dyer procedure55,56. Extracts were evaporated under a stream of nitrogen, re-dissolved in n-hexane and analysed on a Thermo Finnigan Trace GC-FID equipped with a Restek Rxi-5ms capillary column (30 m × 0.25 mm ID). For each 5-cm piece, a ratio between the ({{rm{U}}}_{37}^{{{rm{K}}}^{{prime} }}) values obtained by GC flame ionization detector analysis and MSI of the whole piece was calculated. The average ratio of all pieces for which GC-based values could be obtained was 1.194, with a standard deviation of 0.021.$${{rm{U}}}_{37,{rm{G}}{rm{C}}-{rm{F}}{rm{I}}{rm{D}}}^{{{rm{K}}}^{{prime} }}=1.194times {{rm{U}}}_{37,{rm{M}}{rm{S}}{rm{I}}}^{{{rm{K}}}^{{prime} }}$$
(2)
This ratio was used to calculate GC-equivalent ({{rm{U}}}_{37}^{{{rm{K}}}^{{prime} }}) values, which were then translated into SST using the BAYSPLINE calibration57. The average standard error of the BAYSPLINE model is 0.049 ({{rm{U}}}_{37}^{{{rm{K}}}^{{prime} }}) units (corresponding to 1.4 °C) for samples with SST below 23.4 °C, but increases at higher values (to up to 4.4 °C)57. This is explained by the fact that sensitivity of the ({{rm{U}}}_{37}^{{{rm{K}}}^{{prime} }}) to SST (that is, the slope of the regression) declines at higher values. In the current dataset, the 95% confidence interval is, on average, ±3.6 °C. The analytical precision of MSI-based SST reconstructions for the ({{rm{U}}}_{37}^{{{rm{K}}}^{{prime} }}), using at least ten data points, according to Alfken et al.9, is about 0.3 °C. Sources of uncertainty are summarized in Extended Data Fig. 10a.For frequency analysis, a continuous, annually spaced record was constructed by linearly interpolating 49 missing values. The record was subsequently detrended. Spectral analysis was performed with the REDFIT module58 using a Hanning window (oversample 2, segments 2). Continuous wavelet transforms were applied to investigate changes in cyclicity over time, using the Morlet wavelet with code provided by Torrence and Compo59 for MATLAB. All steps, except for the wavelet analysis, were performed with the PAST software60.For the assessment of the interannual variability, the SST record was band-pass-filtered for periods between 2 and 8 years. The record is based on 1-year binned data; seasonality is thereby nullified and the highest frequency to be evaluated (Nyquist frequency) corresponds to a period of 2 years. Variability of this time series was quantified by calculating the standard deviation of the band-pass-filtered ({{rm{U}}}_{37}^{{{rm{K}}}^{{prime} }}) signal in 25-year intervals. To account for the potential impact of analytical precision on the observed signal (Methods, section titled ‘The effect of changing sedimentation rate on reconstructed interannual SST variability during the YD–Holocene transition’), the variability experiment from Alfken et al.9 was revisited. A sediment extract had been sprayed on an ITO slide and analysed by MSI. We then randomly selected n spots and obtained a ({{rm{U}}}_{37}^{{{rm{K}}}^{{prime} }}) value for the summed intensities of these spots. Precision was calculated as the standard deviation of five replicate experiments for n = 1, 5, 10, 20, 30, 40, 50 and 60. Decreasing analytical variability with increasing number of observations was fitted to a curve (R2 = 0.838) described by the equation$${rm{Analytical}},{rm{variability}}=0.0741times {rm{number}},{rm{of}},{{rm{spots}}}^{-0.558}$$
(3)
On the basis of this equation, analytical variability for each horizon could be calculated on the basis of the number of values included (Extended Data Fig. 10b). The mean variability for each 25-year window was then subtracted from the observed variability in the band-pass-filtered signal and the resulting proxy values were translated to SST following the equation by Müller et al.61. Statistical significance of the change in corrected SST variability after 11.66 kyr b2k was assessed with a t-test.Assessment of SST seasonalityFor the assessment of SST seasonality, alkenone intensities from individual spots were binned according to ΔGS, with a bin size of 1 unit. Spots were then separated into the categories upwelling season and non-upwelling season by identifying the threshold ΔGS value that maximized the difference between average SST in the bins above and below it. Furthermore, this value had to fulfill three conditions: (1) be higher (lighter) than the bins with the highest relative abundance of Ca, Ti and Fe, which is indicative of the dark sediments associated to non-upwelling season, (2) be lower (darker) than the bin with highest relative abundance for Si indicative of light sediment associated to the upwelling season and (3) the number of spots categorized as upwelling and non-upwelling had to account for at least 25% of total spots. If criteria 1 and 2 prevented criteria 3 from being fulfilled, a limit of 15% was set. After separating data into these two categories, data were processed separately as described above for the unfiltered dataset and a downcore temporal resolution of 5 years was applied. Seasonality was calculated as the difference between both records and thus represents the difference between 5-year average SST in the non-upwelling and upwelling seasons.Shift in seasonality was fitted to two different ramps with the RAMPFIT software52. An unconstrained approach and a constrained approach (in which the start and end points of the ramp were restricted to the intervals 11.725–11.8 kyr b2k and 11.6–11.675 kyr b2k) were applied. Negative values were excluded from this fitting. The resulting groups of data were compared by a Mann–Whitney rank test.SST seasonality in the modern Cariaco Basin was calculated for the years 1980 to 2020 based on the HadISST dataset62 by dividing monthly data from each year into two groups and searching for the largest difference between the average temperatures of both groups. Each group had to include at least three consecutive months. In 36 out of 41 years, the warm season was defined from May to November or from July to November.Decadal-scale to centennial-scale SST changes during the YD–Holocene transition and in the early HoloceneAnnually reconstructed SST (average SST = 24.3 °C) remains relatively stable during the YD–Holocene transition. At around 11.4 kyr b2k, a warming trend is observed. Averaging all data before 11.39 kyr and after 11.37 kyr results in a warming from 23.9 ± 1.6 °C to 25.5 ± 1.4 °C. Trends identified by MSI are consistent with conventional ({{rm{U}}}_{37}^{{{rm{K}}}^{{prime} }}) analyses performed in the present study and those previously reported by Herbert and Schuffert23 on Ocean Drilling Program core 165-1002C (Extended Data Fig. 1). These authors observed a slight warming several hundred years after the transition into the Holocene, between about 11.53 and 11.32 kyr b2k.Three prominent SST maxima are observed between about 11.50 and 11.45 kyr b2k. The average SST in these 50 years is 1.3 °C higher than in the 50 years before and after. These maxima are synchronous with the 11.4-ka cold event or PBO characterized by a negative excursion in δ18O and reduced snow accumulation rates in Greenland ice cores63 (Extended Data Fig. 2). The PBO coincides with the oldest of the Bond events, that is, pulses of ice rafting in the Northern Atlantic indicative of climatic deterioration64.A warm tropical response to the PBO would be supported by the lower-resolution foraminiferal SST record of Lea et al.3, which shows two data points of increased SST shortly after the end of the YD–Holocene transition. To enable direct comparison, ages in Lea et al.3 were corrected for the age difference between the sediment-colour-based YD termination midpoint in their record and in data from Deplazes et al.11. After this correction, these maxima correspond to 11.43 and 11.50 kyr b2k (Extended Data Fig. 2). Further, the SST maxima coincide with a short-lived change to lighter-coloured sediments. Hughen et al.19 described a correlation between brief North Atlantic cold events, such as the PBO, and changes in tropical primary productivity mediated by stronger upwelling that result in lighter sediments in the Cariaco Basin. Far-reaching effects of the PBO have previously been described in West Asia, with increased dust plumes being related to a southward shift of the westerlies65.The identification of the mechanisms behind a potential TNA response to the PBO is beyond the scope of this study. However, we wish to point out that high-resolution records are crucial to identify such events and to differentiate between underlying changes coinciding in time and, as in the present case, sharp signals that act on the same multidecadal timescales and can potentially be triggered by the same processes66.The effect of changing sedimentation rate on reconstructed interannual SST variability during the YD–Holocene transitionPooling proxy data into 1-year horizons establishes a constant sampling rate and thereby prevents potential effects of changing sedimentation rates. The onset of the Holocene in the Cariaco Basin sediments is characterized by a sharp decrease in sedimentation rates from 1.4 to 0.5 mm year−1 (refs. 11,19). Consequently, in the yearly pooled data, we observe a reduction in the number of values summed for each horizon (Extended Data Fig. 10b), as fewer laser spots fit into the thinner Holocene annual layers. At the same time, the mean intensity in each of these spots slightly increases, consistent with a relative increase of the contribution of haptophytes to primary production20.We have previously shown that the precision of MSI-based molecular proxy analysis is dependent on both the number of spots pooled per data point and the signal intensity in these spots54. All horizons used in the downcore record are above the established threshold of ten spots and proxy variability was shown to stabilize above this threshold9,54. However, as a decrease in the number of values per horizon might still result in lower analytical precision and contribute to higher signal variability, we corrected variability in the 2–8-year window with the estimated analytical variability (see equation (3)). With this correction, the magnitude of the described variability decreases across the record, but the trend towards higher interannual variability in the Holocene persists (Fig. 2c).Varve formation and alkenone deposition in the sediments of the Cariaco Basin during the YD–Holocene transitionComparison of elemental maps and sediment colour (Extended Data Fig. 5) shows a consistent pattern of lamination across the YD–Holocene transition that results from the seasonal interplay of precipitation, upwelling and dominant phytoplankton community composition. Darker laminae represent the rainy, non-upwelling (summer/fall) season and are enriched in Fe and Ti from terrigenous material and Ca sourced from biogenic CaCO3 produced by foraminifera or coccolithophores. Lighter laminae are characterized by high abundance of Si and correspond to the increased production of biogenic opal by diatoms during the upwelling (winter/spring) season67. This is in agreement with observations by Hughen et al.18, who described the laminae couplets in the Cariaco Basin as representing annual cycles, whereby light laminae are an indicator of high productivity associated with the winter/spring upwelling season and dark laminae are an indicator of summer/fall runoff and accumulation of terrigenous material. Deplazes et al.68 described a divergent origin of lamination for a deeper section of the YD, with light laminae being rich in calcareous and terrigenous elements characteristic for the summer season, whereas dark layers were enriched in Si and Br, indicative of diatoms and organic-walled primary producers characteristic for the more productive winter season. Such an alteration of the characteristic pattern of lamination is not observed in the late YD investigated here.This blueprint of seasonality was used to assess the seasonal behaviour of alkenones. Alkenones were deposited throughout the year, as evidenced by the fact that the number of spots containing detectable amounts of both alkenone species are not restricted to the upwelling or non-upwelling seasons but distributed across a relatively wide range of GS values to both sides of the median (Extended Data Fig. 6). Average alkenone signal intensity is higher in the non-upwelling season, pointing to a preference of alkenone producers for this season and/or to a stronger dilution of the signal in the upwelling season. In regards to the ({{rm{U}}}_{37}^{{{rm{K}}}^{{prime} }}) SST proxy distribution in light versus dark layers, our observations are in agreement with the ability to capture the seasonal SST cycle with alkenones in sinking particles in the modern Cariaco Basin69.Effect of changing seasonality on YD and early Holocene SST records from the western TNAChanging seasonality can contribute to explaining contrasting lower-resolution SST records in the western TNA during the YD and the early Holocene. The strong warming during the YD–Holocene transition recorded in the foraminiferal Mg/Ca record of the Cariaco Basin (Lea et al.3; Extended Data Fig. 1) might be reflecting the more robust thermohaline stratification and increasingly warmer non-upwelling seasons, given the preference of Globigerinoides ruber for this season.Globigerinoides ruber (white), as used by Lea et al.3, is considered to be a dominant species in the tropics, with a relatively uniform annual distribution. However, in the modern Cariaco Basin, upwelling leads to a distinct foraminiferal community composition and seasonal turnover70, consistent with the notion of warm-water foraminifera narrowing their occurrence to the warmest season71. The relative abundance of G. ruber increases in the non-upwelling (warm) season but rarely exceeds 15%, whereas the upwelling season is clearly dominated by Globigerina bulloides72,73. Globigerinoides ruber fluxes are consistently lowest when upwelling is most vigorous, as expressed in annual minima in SST (Extended Data Fig. 9b). As upwelling during the YD and early Holocene was more intense than in the present70, the preference of G. ruber for the summer (non-upwelling) season might have been even more pronounced.The development of a stronger seasonality in the early Holocene would thus have led to a narrower temporal occurrence of G. ruber in the non-upwelling season, during which it would also be exposed to higher SST. The average SST difference between seasons obtained in our analysis can be converted into annual SST amplitude by assuming a sinusoidal curve. By doing so, we observe an increase in the seasonal amplitude of 1.5 to 1.9 °C (depending on the ramp fitted), which is similar to the warming described by Lea et al.3.This interpretation is in agreement with Bova et al.46, who observed that most Holocene climate reconstructions are biased towards the boreal summer/fall and reflect the evolution of seasonal rather than annual temperatures. As discussed above, this is probably not true for the ({{rm{U}}}_{37}^{{{rm{K}}}^{{prime} }}) index in the Cariaco Basin, as alkenones are deposited throughout the year. The suggested weakening of summer stratification during the YD (as compared with the Holocene) might, however, explain why the lower-resolution ({{rm{U}}}_{37}^{{{rm{K}}}^{{prime} }}) records from the semi-enclosed Cariaco Basin show no or weaker warming23 than other, open-ocean, tropical YD records4, where the interplay of upwelling, freshwater input and stratification are less relevant to the SST signal. More