Philippa Kaur

Rasmussen, S. O. et al. A stratigraphic framework for abrupt climatic changes during the Last Glacial period based on three synchronized Greenland ice-core records: refining and extending the INTIMATE event stratigraphy. Quat. Sci. Rev. 106, 14–28 (2014).Article 

Google Scholar 
Heiri, O. et al. Validation of climate model-inferred regional temperature change for late-glacial Europe. Nat. Commun. 5, 1–7 (2014).Article 
CAS 

Google Scholar 
Muschitiello, F. et al. Fennoscandian freshwater control on Greenland hydroclimate shifts at the onset of the Younger Dryas. Nat. Commun. 6, 1–8 (2015).Article 
CAS 

Google Scholar 
Renssen, H. et al. Multiple causes of the Younger Dryas cold period. Nat. Geosci. 8, 946–949 (2015).CAS 
Article 

Google Scholar 
Mangerud, J. The discovery of the Younger Dryas, and comments on the current meaning and usage of the term. Boreas 50, 1–5 (2021).Article 

Google Scholar 
Cheng, H. et al. Timing and structure of the Younger Dryas event and its underlying climate dynamics. Proc. Natl. Acad. Sci. USA 117, 23408–23417 (2020).CAS 
Article 

Google Scholar 
van Hoesel, A., Hoek, W. Z., Pennock, G. M. & Drury, M. R. The Younger Dryas impact hypothesis: a critical review. Quat. Sci. Rev. 83, 95–114 (2014).Article 

Google Scholar 
Partin, J. W. et al. Gradual onset and recovery of the Younger Dryas abrupt climate event in the tropics. Nat. Commun. 6, 1–9 (2015).Article 

Google Scholar 
Reinig, F. et al. Precise date for the Laacher See eruption synchronizes the Younger Dryas. Nature 595, 66–69 (2021).CAS 
Article 

Google Scholar 
Ammann, B. et al. Quantification of biotic responses to rapid climatic changes around the Younger Dryas — a synthesis. Palaeogeogr. Palaeoclimatol. Palaeoecol. 159, 313–347 (2000).Article 

Google Scholar 
Hoek, W. Z. Vegetation response to the ∼ 14.7 and ∼ 11.5 ka cal. BP climate transitions: is vegetation lagging climate? Glob. Planet. Change 30, 103–115 (2001).Article 

Google Scholar 
Litt, T. et al. Correlation and synchronisation of Lateglacial continental sequences in northern central Europe based on annually laminated lacustrine sediments. Quat. Sci. Rev. 20, 1233–1249 (2001).Article 

Google Scholar 
Muschitiello, F. & Wohlfarth, B. Time-transgressive environmental shifts across Northern Europe at the onset of the Younger Dryas. Quat. Sci. Rev. 109, 49–56 (2015).Article 

Google Scholar 
Nakagawa, T. et al. The spatio-temporal structure of the Lateglacial to early Holocene transition reconstructed from the pollen record of Lake Suigetsu and its precise correlation with other key global archives: implications for palaeoclimatology and archaeology. Glob. Planet. Change 202, 103493 (2021).Article 

Google Scholar 
Ammann, B. et al. Vegetation responses to rapid warming and to minor climatic fluctuations during the Late-Glacial Interstadial (GI-1) at Gerzensee (Switzerland). Palaeogeogr. Palaeoclimatol. Palaeoecol. 391, 40–59 (2013).Article 

Google Scholar 
Engels, S. et al. Subdecadal‐scale vegetation responses to a previously unknown late‐Allerød climate fluctuation and Younger Dryas cooling at Lake Meerfelder Maar (Germany). J. Quat. Sci 31, 741–752 (2016).Article 

Google Scholar 
Van Raden, U. J. et al. High-resolution late-glacial chronology for the Gerzensee lake record (Switzerland): δ18O correlation between a Gerzensee-stack and NGRIP. Palaeogeogr. Palaeoclimatol. Palaeoecol. 391, 13–24 (2013).Article 

Google Scholar 
Blaga, C. I., Reichart, G.-J., Lotter, A. F., Anselmetti, F. S. & Sinninghe Damsté, J. S. A TEX86 lake record suggests simultaneous shifts in temperature in Central Europe and Greenland during the last deglaciation. Geophys. Res. Lett. 40, 948–953 (2013).Article 

Google Scholar 
Rach, O., Brauer, A., Wilkes, H. & Sachse, D. Delayed hydrological response to Greenland cooling at the onset of the Younger Dryas in western Europe. Nat. Geosci. 7, 109 (2014).CAS 
Article 

Google Scholar 
Strogatz, S. H. Exploring complex networks. Nature 410, 268–276 (2001).CAS 
Article 

Google Scholar 
Doncaster, C. P. et al. Early warning of critical transitions in biodiversity from compositional disorder. Ecology 97, 3079–3090 (2016).Article 

Google Scholar 
Jones, G. et al. The Lateglacial to early Holocene tephrochronological record from Lake Hämelsee, Germany: a key site within the European tephra framework. Boreas 47, 28–40 (2018).Article 

Google Scholar 
Blaga, C. I., Reichart, G.-J., Heiri, O. & Damsté, J. S. S. Tetraether membrane lipid distributions in water-column particulate matter and sediments: a study of 47 European lakes along a north–south transect. J. Paleolimnol. 41, 523–540 (2009).Article 

Google Scholar 
Bechtel, A., Smittenberg, R. H., Bernasconi, S. M. & Schubert, C. J. Distribution of branched and isoprenoid tetraether lipids in an oligotrophic and a eutrophic Swiss lake: insights into sources and GDGT-based proxies. Org. Geochem. 41, 822–832 (2010).CAS 
Article 

Google Scholar 
Lowe, J. et al. On the timing of retreat of the Loch Lomond (‘Younger Dryas’) Readvance icefield in the SW Scottish Highlands and its wider significance. Quat. Sci. Rev. 219, 171–186 (2019).Article 

Google Scholar 
Muggeo, V. M. R. Segmented: an R package to fit regression models with broken-line relationships. R news 8, 20–25 (2008).
Google Scholar 
Merkt, J. & Müller, H. Varve chronology and palynology of the Lateglacial in Northwest Germany from lacustrine sediments of Hämelsee in Lower Saxony. Quat. Int. 61, 41–59 (1999).Article 

Google Scholar 
Litt, T. & Stebich, M. Bio-and chronostratigraphy of the lateglacial in the Eifel region, Germany. Quat. Int. 61, 5–16 (1999).Article 

Google Scholar 
Reimer, P. J. et al. The IntCal20 Northern hemisphere radiocarbon age calibration curve (0–55 cal kBP). Radiocarbon 62, 725–757 (2020).CAS 
Article 

Google Scholar 
Giesecke, T. Holocene dynamics of the southern boreal forest in Sweden. The Holocene 15, 858–872 (2005).Article 

Google Scholar 
Müller, D. et al. New insights into lake responses to rapid climate change: the Younger Dryas in Lake Gościąż, central Poland. Boreas 50, 535–555 (2021).Article 

Google Scholar 
Davis, B. A. S. et al. The Eurasian Modern Pollen Database (EMPD), version 2. Earth Syst. Sci. data 12, 2423–2445 (2020).Article 

Google Scholar 
Neugebauer, I. et al. A Younger Dryas varve chronology from the Rehwiese palaeolake record in NE-Germany. Quat. Sci. Rev. 36, 91–102 (2012).Article 

Google Scholar 
Ralska-Jasiewiczowa, M. et al. Very fast environmental changes at the Pleistocene/Holocene boundary, recorded in laminated sediments of Lake Gościaż, Poland. Palaeogeogr. Palaeoclimatol. Palaeoecol. 193, 225–247 (2003).Article 

Google Scholar 
Bonk, A. et al. Varve microfacies and chronology from a new sediment record of Lake Gościąż (Poland). Quat. Sci. Rev. 251, 106715 (2021).Article 

Google Scholar 
Brauer, A., Haug, G. H., Dulski, P., Sigman, D. M. & Negendank, J. F. W. An abrupt wind shift in western Europe at the onset of the Younger Dryas cold period. Nat. Geosci. 1, 520–523 (2008).CAS 
Article 

Google Scholar 
Mekhaldi, F. et al. Radionuclide wiggle matching reveals a nonsynchronous early Holocene climate oscillation in Greenland and western Europe around a grand solar minimum. Clim. Past 16, 1145–1157 (2020).Article 

Google Scholar 
Mayfield, R. J. et al. Metrics of structural change as indicators of chironomid community stability in high latitude lakes. Quat. Sci. Rev. 249, 106594 (2020).Article 

Google Scholar 
van der Knaap, W. O. & Van Leeuwen, J. F. N. Climate-pollen relationships AD 1901–1996 in two small mires near the forest limit in the northern and central Swiss Alps. The Holocene 13, 809–828 (2003).Article 

Google Scholar 
Bazelmans, J. et al. Environmental changes in the late Allerød and early Younger Dryas in the Netherlands: a multiproxy high-resolution record from a site with two Pinus sylvestris populations. Quat. Sci. Rev. 272, 107199 (2021).Article 

Google Scholar 
Birks, H. H., Battarbee, R. W. & Birks, H. J. B. The development of the aquatic ecosystem at Kråkenes Lake, western Norway, during the late glacial and early Holocene-a synthesis. J. Paleolimnol 23, 91–114 (2000).Article 

Google Scholar 
Bronk Ramsey, C. Bayesian analysis of radiocarbon dates. Radiocarbon 51, 337–360 (2009).Article 

Google Scholar 
Lohne, Ø. S., Mangerud, J. A. N. & Birks, H. H. IntCal13 calibrated ages of the Vedde and Saksunarvatn ashes and the Younger Dryas boundaries from Kråkenes, western Norway. J. Quat. Sci 29, 506–507 (2014).Article 

Google Scholar 
Lohne, Ø. S., Mangerud, J. A. N. & Birks, H. H. Precise 14 C ages of the Vedde and Saksunarvatn ashes and the Younger Dryas boundaries from western Norway and their comparison with the Greenland Ice Core (GICC 05) chronology. J. Quat. Sci 28, 490–500 (2013).Article 

Google Scholar 
Wohlfarth, B. et al. Hässeldala–a key site for last termination climate events in northern Europe. Boreas 46, 143–161 (2017).Article 

Google Scholar 
Brauer, A. et al. High resolution sediment and vegetation responses to Younger Dryas climate change in varved lake sediments from Meerfelder Maar, Germany. Quat. Sci. Rev. 18, 321–329 (1999).Article 

Google Scholar 
Lane, C. S., Brauer, A., Blockley, S. P. E. & Dulski, P. Volcanic ash reveals time-transgressive abrupt climate change during the Younger Dryas. Geology 41, 1251–1254 (2013).Bronk Ramsey, C. et al. Improved age estimates for key Late Quaternary European tephra horizons in the RESET lattice. Quat. Sci. Rev. 118, 18–32 (2015).Rasmussen, S. O. et al. A new Greenland ice core chronology for the last glacial termination. J. Geophys. Res. Atmos. 111, https://doi.org/10.1029/2005JD006079 (2006).Brauer, A., Endres, C., Zolitschka, B. & Negendank, J. F. W. AMS radiocarbon and varve chronology from the annually laminated sediment record of Lake Meerfelder Maar, Germany. Radiocarbon 42, 355–368 (2000).CAS 
Article 

Google Scholar 
Wulf, S. et al. Tracing the Laacher See Tephra in the varved sediment record of the Trzechowskie palaeolake in central Northern Poland. Quat. Sci. Rev. 76, 129–139 (2013).Article 

Google Scholar 
Brauer, A. et al. The importance of independent chronology in integrating records of past climate change for the 60–8 ka INTIMATE time interval. Quat. Sci. Rev. 106, 47–66 (2014).Article 

Google Scholar 
Lane, C. S. et al. The Late Quaternary tephrostratigraphy of annually laminated sediments from Meerfelder Maar, Germany. Quat. Sci. Rev. 122 192–206 (2015).Article 

Google Scholar 
Adolphi, F. & Muscheler, R. Synchronizing the Greenland ice core and radiocarbon timescales over the Holocene–Bayesian wiggle-matching of cosmogenic radionuclide records. Clim. Past 12, 15–30 (2016).Article 

Google Scholar 
Muschitiello, F. et al. Deep-water circulation changes lead North Atlantic climate during deglaciation. Nat. Commun. 10, 1–10 (2019).CAS 
Article 

Google Scholar 
Adolphi, F. et al. Persistent link between solar activity and Greenland climate during the Last Glacial Maximum. Nat. Geosci. 7, 662–666 (2014).CAS 
Article 

Google Scholar 
Siegenthaler, U., Heimann, M. & Oeschger, H. 14C variations caused by changes in the global carbon cycle. Radiocarbon 22, 177–191 (1980).CAS 
Article 

Google Scholar 
Muscheler, R., Adolphi, F. & Svensson, A. Challenges in 14C dating towards the limit of the method inferred from anchoring a floating tree ring radiocarbon chronology to ice core records around the Laschamp geomagnetic field minimum. Earth Planet. Sci. Lett. 394, 209–215 (2014).CAS 
Article 

Google Scholar 
Muschitiello, F. An improved and continuous synchronization of the Greenland ice-core and Hulu Cave U-Th timescales using probabilistic inversion. Clim. Past Discuss. 1–39 https://doi.org/10.5194/cp-2021-116 (2021).Moore, P. D., Webb, J. A. & Collison, M. E. Pollen analysis. (Blackwell scientific publications, 1991).Engels, S. et al. Haemelsee: late-glacial pollen counts. PANGAEA, https://doi.org/10.1594/PANGAEA.939693 (2021).Weltje, G. J. & Tjallingii, R. Calibration of XRF core scanners for quantitative geochemical logging of sediment cores: Theory and application. Earth Planet. Sci. Lett. 274, 423–438 (2008).CAS 
Article 

Google Scholar 
Heiri, O., Lotter, A. F. & Lemcke, G. Loss on ignition as a method for estimating organic and carbonate content in sediments: reproducibility and comparability of results. J. Paleolimnol 25, 101–110 (2001).Article 

Google Scholar 
Brooks, S. J., Langdon, P. G. & Heiri, O. The identification and use of Palaearctic Chironomidae larvae in palaeoecology. Quat. Res. Assoc. Tech. Guid. i–vi. Vol. 10, 1–276 (2007).Heiri, O., Brooks, S. J., Birks, H. J. B. & Lotter, A. F. A 274-lake calibration data-set and inference model for chironomid-based summer air temperature reconstruction in Europe. Quat. Sci. Rev. 30, 3445–3456 (2011).Article 

Google Scholar 
Heiri, O. & Lotter, A. F. Effect of low count sums on quantitative environmental reconstructions: an example using subfossil chironomids. J. Paleolimnol 26, 343–350 (2001).Article 

Google Scholar 
Rach, O., Hadeen, X. & Sachse, D. An automated solid phase extraction procedure for lipid biomarker purification and stable isotope analysis. Org. Geochem. 142, 103995 (2020).CAS 
Article 

Google Scholar 
Huguet, C. et al. An improved method to determine the absolute abundance of glycerol dibiphytanyl glycerol tetraether lipids. Org. Geochem. 37, 1036–1041 (2006).CAS 
Article 

Google Scholar 
Hopmans, E. C., Schouten, S. & Damsté, J. S. S. The effect of improved chromatography on GDGT-based palaeoproxies. Org. Geochem. 93, 1–6 (2016).CAS 
Article 

Google Scholar 
Birks, H. J. B. & Birks, H. H. Biological responses to rapid climate change at the Younger Dryas—Holocene transition at Kråkenes, western Norway. The Holocene 18, 19–30 (2008).Article 

Google Scholar 
R CORE TEAM, A. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. 2012. URL http://www.R-project.org (2020).Engels, S., van Geel, B., Buddelmeijer, N. & Brauer, A. High-resolution palynological evidence for vegetation response to the Laacher See eruption from the varved record of Meerfelder Maar (Germany) and other central European records. Rev. Palaeobot. Palynol. 221, 160–170 (2015).Article 

Google Scholar 
Hughes, A. L. C., Gyllencreutz, R., Lohne, Ø. S., Mangerud, J. & Svendsen, J. I. The last Eurasian ice sheets–a chronological database and time‐slice reconstruction, DATED‐1. Boreas 45, 1–45 (2016).Article 

Google Scholar  More

The CAN-SAR22 database was created to provide access to publicly available data on species at risk in Canada in a standardized format that can be used in a wide range of applied research contexts. The variables included in the database were chosen to provide a range of information available for species at risk with a particular focus on climate change to support the first publication using the database6. The database includes numerous data fields including extinction risk status, various biological and geographical attributes, threat assessments, date of listing, recovery actions, and a set of climate change impact and adaptation variables. CAN-SAR is a living database that can be updated as new information and reports become available, or as other targeted data extraction efforts become available23.In Canada, the listing process begins with an assessment of a wildlife species’ risk of extinction by the Committee on the Status of Endangered Wildlife in Canada (COSEWIC). A wildlife species can be either a species or a ‘designatable unit’, which includes subspecies, varieties, or other geographically or genetically distinct populations. Herein these are referred to collectively as ‘species’. COSEWIC is an independent body of experts who synthesize the best available information to date into a status report containing elements such as population size and trends, habitat availability, and threat assessments (Fig. 1)17. This report is then used as the basis for a status recommendation that is passed on to the Government of Canada, who makes the final decision on whether to legally list the species under Schedule 1 of SARA24. The species can be listed as ‘Special concern’, ‘Threatened’, ‘Endangered’, or ‘Extirpated’. If a species is listed as ‘Threatened’, ‘Endangered’ or ‘Extirpated’ then a recovery strategy is required, while for species listed as ‘Special concern’ a management plan must be created24. Recovery strategies must provide a description of the species’ needs, address identified threats, identify critical habitat (where applicable and to the extent possible), and include population and distribution objectives for the species’ recovery. Management plans include conservation measures for the species and its habitat24. Hereafter, we refer to recovery strategies and management plans collectively as ‘recovery documents’.Information included in the database was extracted from various sources and documents that are available from the online SAR Public Registry, including COSEWIC status reports and status appraisal summaries, and recovery documents (Fig. 1). A COSEWIC status appraisal summary is produced instead of a new status report when a species has been previously assessed and COSEWIC experts are confident that its status will not change (https://www.cosewic.ca/index.php/en-ca/assessment-process/status-appraisal-summary-process.html). It is considered an addendum to the existing status report; thus, we use ‘status report’ to refer to either a status report or a status appraisal summary and the previous status report. From the SAR Public Registry website we accessed information from 1146 documents for all 594 species listed under SARA Schedule 1 as of March 23, 2021, that were classified with the status of ‘Special concern’, ‘Threatened’, or ‘Endangered’. Some species have multiple documents of the same type because COSEWIC reassesses at risk species every 10 years or less and recovery strategies and management plans are reviewed every 5 years and updated as needed. As new documents have become available they have been added to the CAN-SAR database without overriding the previously existing document, which allows for tracking of changes in various data fields over time. Only documents between 2018 and 2021, inclusive, have an updated version due to our updating schedule.Data extractionVariables included in the CAN-SAR database were categorised as either directly transcribed or derived. Directly transcribed variables reflect information extracted from documents that require limited interpretation, such as scientific name or date of legal listing (Online-only Table 1). Derived variables reflect species’ attributes that required interpretation of text by data recorders (Online-only Table 1). The data dictionary (CAN-SAR_data_dictionary.xlsx) contains a description of each variable, including details of their extraction and synthesis22.Several derived variables were extracted from the status report technical summary section, including whether the species is endemic to Canada or North America, and whether the species’ range is continuous with the United States. Endemism was determined for each species at two spatial extents, Canada and North America, based on descriptions of their global distributions from status reports. Whether a Canadian species’ range is continuous with its conspecifics in the United States was interpreted from descriptions of geographic isolation in the distribution and rescue effect sections of the status reports.Variables related to species’ threats were derived from information in the status reports, recovery strategies and management plans. In 2012, COSEWIC initiated use of the IUCN threats classification system in status reports for some species; a ‘threats calculator’25. Threats calculators may also be included in recovery strategies and management plans. A threats calculator is a table included in the document that classifies threats into 11 general ‘level one’ classes and, more specific ‘level two’ subclasses (Table 1)26. Four variables (impact, severity, scope, and timing) for each level one and level two threats were scored independently and then combined into an overall impact score for each species. Impact is defined as the degree to which the species is threatened by the threat class; severity is the level of damage to the species from the threat class that is expected within ten years or three generations, whichever is longer; scope is the proportion of the species that is expected to be affected within ten years; and timing is the immediacy of the threat25. Threat-related variables were either transcribed directly from the threats calculator, or from the derived description of threats in the document if a threats calculator was not included.Table 1 Definitions of level one threat classes and names of level two threat classes following Version 1.1 of the IUCN threats classification system.Full size tableFor species where a threats calculator was included, we recorded whether each of the level one and level two threat classes were identified (i.e., considered a threat), and transcribed the scores for each of impact, scope, severity, and timing. Threat classes were considered identified if the impact was negligible, low, moderate, high, very high, unknown, or not calculated (outside assessment timeframe). Impact, scope, severity, and timing values were coded as ranked values of ‘0’: not a threat; ‘1’: neglible; ‘2’: low; ‘3’: moderate; ‘4’: high; ‘5’: very high; ‘-1’: unknown; ‘-2’: not calculated; or ‘NA’ where there were blank values. For exact ranking interpretations see CAN-SAR_data_dictionary22. For some species, the threats calculator was available from the COSEWIC Secretariat as a Microsoft Excel file, in which case threats information was extracted directly from the spreadsheet using R v 3.6.227. For species where a Microsoft Excel file was not available, threats calculator information was manually extracted from the status report.For species where a threats calculator was not included in the document, threats described in the text were classified into threat classes based on version 1.1 of the IUCN threats classification system (Table 1)26. Although a more recent version of the threats calculator exists, we applied version 1.1 classification to reflect the approach applied across the majority of species. Threats were considered identified if the threat was discussed as having any negative or potentially negative impact on the species. In cases where no threat calculator was available, the threat attributes of impact, scope, severity, and timing were scored as not applicable; ‘NA’.Several variables were derived to determine how climate change was addressed in status reports and recovery documents. Whether climate change was mentioned anywhere in the status report was determined by searching the document for the words climat*, warm, temperat*, and drought. If a document contained any of these search terms, we assessed the context for description of anthropogenic climate change impacts. In cases where the terms were not found, the threats section was checked for any other descriptions that were related to climate change; if none were found, climate change was recorded as not mentioned. When climate change was mentioned, we then determined if it was identified as a threat by interpreting whether it was described as having a negative or potentially negative impact on the species. If a threats calculator was included in the status report, climate change was considered a threat if the ‘Climate change and severe weather’ threat class had an impact that was more than negligible or if climate change was described outside the threats calculator as a threat or potential threat. We recorded whether the threat of climate change was unknown. This included instances where climate change was described as having unknown effects on the species, if ‘unknown’ was assigned to impact, scope, severity, or timing in the threats calculator, or if knowledge gaps related to climate change were identified. Finally, the impact of climate change relative to other threats was classified based on descriptions of threats in the status report. The relative impact of climate change was classified as ‘0’ if it was not a threat; ‘1’ if it was described as a minor, potential, possible, or other threat; ‘2’ if it was a significant threat but not the most important or if it was among the list of threats with no indication of relative importance; or ‘3’ if it was among the most important threats described.Additional derived variables extracted from recovery documents available on the SAR Public Registry included those related to critical habitat identification and recovery actions. For species with recovery strategies, we recorded whether critical habitat was described as identified, partially identified, or not identified. In cases where critical habitat was described as “identified to the extent possible”, it was marked as identified. We extracted information from recovery documents on what types of actions were recommended and whether the actions addressed the threat of climate change. Actions were categorized into four categories: outreach and stewardship, research and monitoring, habitat management, and population management (Table 2). Within each of the four categories, a set of 16 sub-types were recorded if any actions of that type were recommended or already completed. We also recorded action types and sub-types that specifically addressed climate change threats if climate change was listed as the threat addressed or the reason the action was necessary6.Table 2 Categories of actions specified in Recovery Strategies.Full size tableFive data recorders conducted the initial data extraction, synthesis, and interpretation. All recorders were trained on the definitions, interpretation, and general process of data extraction to ensure consistent extraction of all variables. Data extraction occurred in multiple stages and included an iterative set of verifications and assessments of the same species among recorders to ensure consistent and standardized interpretations. Once convergence of interpretations was achieved, each recorder was assigned a set of species/reports from which to extract information.Next stepsThe CAN-SAR database is intended to be a living database that can be updated by adding information from new documents or species as they become available, adding more historical documents, or extracting new information from all documents. The current set of species and associated information includes those listed on Schedule 1 of SARA (as of March 23rd 2021) as ‘Special concern’, ‘Threatened’, or ‘Endangered’. Examples of future data additions include integration of data from species assessed by COSEWIC that are not listed under Schedule 1 of SARA, adding fields that specify the criteria used to arrive at a risk status designation, and integration of data from action plans. We anticipate updating the database periodically, as time and resources allow, and we also encourage anyone interested in extending or expanding on the CAN-SAR database to communicate to discuss a collaboration. Integration of new datasets will require screening and validation to ensure adherence to data standards and consistent interpretations. In the longer term, we foresee the implementation of automatic updating of the CAN-SAR database for variables that do not require interpretation by using machine-readable formatted status and recovery documents.ApplicationsApplications of the CAN-SAR database reflect both opportunities to synthesise the data in novel ways and to expand the scope of the current database to include new data fields representing information contained in status assessments and recovery documents. The CAN-SAR database facilitates independent data analysis and synthesis efforts ranging from trend analysis of threats, identifying research and monitoring gaps, and assessing the effectiveness of recovery actions, which target various steps of the listing and recovery process. For example, the database provides a platform to extend existing climate change focused work6 to assess the prevalence of recommended climate change targeted recovery actions, such as translocations. With recent adoption of the ‘Pan-Canadian approach to transforming Species at Risk conservation in Canada’28, which emphasizes multi-species recovery planning approaches, there is an opportunity to assess patterns in key sectors, which include agriculture, forestry, and urban development, over time and by taxa and how they map to threats.With the integration of additional variables through future data extraction or integration efforts, the CAN-SAR database can be used to assess novel questions. For example, broadening recovery action categories to include those that reflect natural climate solutions can highlight where recovery efforts may provide co-benefits, thus achieving biodiversity conservation and climate change mitigation goals29. Specifically, habitat restoration actions for a forest-dependent species primarily threatened by habitat loss may lead to improved recovery outcomes while also resulting in carbon sequestration and improved climate change mitigation efforts. Tracking these types of actions in CAN-SAR could highlight both opportunities and gaps for the integration of climate smart conservation principles30 into species at risk recovery planning and the adoption of climate change adaption measures for species directly considered climate change threatened and those that are not6. More

cpDNA haplotype databaseDNA sequencing of the choloroplast (cp) markers produced sequences of the following lengths: 573 bp (atpB-rbcL); 487 bp (petG-trnP); 500 bp (trnL1-trnL2); and 593 bp (psbM-trnD). Alignment of the 352 individuals from the 44 populations yielded a total 28 variable sites: 11 in the atpB-rbcL spacer, seven in both the petG-trnP and psbM-trnD spacers, and three in the trnL1-trnL2 spacer (Supplementary Table S1). Based on these 28 variable sites (21 base substitutions and 7 deletions) across the combined intergenic regions, a total of 22 unique haplotypes were found (Fig. 1a).Figure 1(a) Chloroplast haplotype distribution in the Shorea leprosula populations. The pie chart colours indicate haplotype distributions; and sector areas are proportional to sample size (Map was generated by ArcGIS-ArcMap version 10.8). (b) STRUCTURE analysis identified two clusters (K = 2) corresponding to Region A and B.Full size imageSSR allele frequency databaseThe reproducibility of SSR genotyping was confirmed by achieving consistent genotypes from five independent PCR amplifications on a single individual for each of the ten SSR loci. Individual bar plots from STRUCTURE analysis are presented in Fig. 1b. At the highest Delta K likelihood scores, the best representation of the data was K = 2 suggesting that the 44 populations in Peninsular Malaysia can be divided into two main genetic clusters: Region A and Region B. The first cluster, ‘Region A’ consists of 12 populations, namely SBadak, BPerangin, BEnggang, GJerai, RTelui, GInas, GBongsu, Belum, Piah, BHijau, Korbu and Bubu. The second cluster, ‘Region B’ consists of 32 populations, namely Behrang, Ampang, HGombak, HLangat, SLalang, PPanjang, Berembun, Angsi, Kenaboi, Triang, Pasoh, BSenggeh, GLedang, Krau, TNegara, Terenggun, SBetis, USat, CTongkat, HTerengganu, Jengai, AGading, Tekam, Beserah, Jengka, Lentang, Lesong, ERompin, GArong, Labis, AHitam and Panti. Similarly, the UPGMA dendrogram analysis also divided the 44 populations into two genetic clusters (Fig. 2) corresponding to Region A and B of the STRUCTURE result.Figure 2Dendrogram showing the relationship between 44 populations of Shorea leprosula in Peninsular Malaysia based on the UPGMA cluster analysis of SSR markers.Full size imageSSR allele frequency databases were established according to Region A and B, and characterized to evaluate the relative usefulness of each SSR marker in forensic investigation. The distribution of allele frequencies for each locus is listed in Table S2 (Region A database) and Table S3 (Region B database). Forensic parameters are shown in Table 1, with a total of 143 alleles and 174 alleles detected in the Region A and B databases, respectively. The observed (Ho) and expected (He) heterozygosity ranged from 0.3570 to 0.8346 and 0.4375 to 0.8795, respectively for populations in the Region A database; and ranged from 0.3298 to 0.8356 and 0.3469 to 0.8793, respectively for populations in the Region B database. The power of discrimination (PD) for the SSR loci ranged from 0.601 to 0.972 and 0.554 to 0.975, in Region A and B databases, respectively. The most discriminating locus was Sle605 in both the Region A (PD = 0.972) and Region B (PD = 0.975) databases. Minimum allele frequency was adjusted for alleles falling below the thresholds of 0.0066 (Region A) and 0.0024 (Region B).Table 1 Genetic diversity and forensic variables (A: total number of alleles; Ho: observed heterozygosity; He: expected heterozygosity; PIC: polymorphic information content; HWE: Hardy–Weinberg equilibrium; MP: matching probability; PD: power of discrimination) for each the 10 SSR loci of Shorea leprosula in the Region A and B databases.Full size tableDeviations from HWE were detected in four of the SSR loci for Region A (SleT11, SleT15, SleT17 and Sle465) and six SSR loci in Region B (SleT01, SleT11, SleT15, SleT17, SleT29 and SleT31). We evaluated these loci in each population independently to rule out the possible presence of null alleles. There were four populations in Region A (GJerai, RTelui, GBongsu and Piah) where a single one locus deviated from HWE; whereas there were eight populations in Region B (Behrang, HGombak, SLalang, Angsi, Klau, USat, Jengka and Panti) with a single locus and a single population (GLedang) with two loci that deviated from HWE (Table S4). Observed deviation from HWE was substantially lower in each population (either absence or not more than two loci) and thus it might be due to Wahlund effect caused by population substructuring in both Region A and B. Linkage disequilibrium (LD) testing was used to evaluate the independence of frequencies for all the SSR genotypes. A total of 13.3% and 28.9% of the 45 pairwise loci were found significant evidence of LD for Region A and B, respectively. Some of the loci might be linked as a result of population substructuring and inbreeding (inbreeding coefficient = 0.0822 [Peninsular Malaysia]). These results are in line with observations in real populations, where the assumption of completely random mating and zero migration required for HWE and LD are unlikely to be met, either in humans, animals or plants 21,22,23.Mean self-assignment, the proportion of individuals correctly assigned back to their population, was 45.9% and ranged from 14.3% (Kenaboi) to 81.3% (CTongkat) between population (Table 2). At the regional level, correct assignment rate of individuals to their region of origin was higher, 87.4% for Region A and 90.0% for Region B, (average of 88.7%).Table 2 Self-assignment test outcomes for Shorea leprosula individuals at the population and regional levels.Full size tableConservativeness of the databaseThe coancestry coefficient (θ) for Peninsular Malaysia (0.0579) was higher than those of Region A (0.0454) and Region B (0.0500) (Table 3). A total of 4.54% and 5.00% of the genetic variability was distributed among populations within Region A and Region B, respectively. In terms of inbreeding coefficient (f), the value for the Region A database (f = 0.0892) was highest, followed by Peninsular Malaysia (f = 0.0822) and Region B (f = 0.0666). All the θ and f values were significantly greater than zero, demonstrated by the 95% confidence intervals not overlapping with zero. Both of the θ and f values were used to calculate the conservativeness of each database by testing the cognate database (Porigin) against the regional database (Pcombined). The databases were non-conservative at the calculated θ value. In order for both the Region databases (A and B) to be conservative, the value of θ was adjusted from 0.0454 to 0.1900 for Region A and from 0.0500 to 0.1500 for Region B. For the Region A database, the most common SSR profile frequency is 2.69 × 10–7 or 1 in 3.72 million and the rarest profile frequency is 1.84 × 10–14 or 1 in 54.3 trillion. For the Region B database, the most common SSR profile frequency is 1.06 × 10–7 or 1 in 9.43 million and the rarest profile frequency is 4.03 × 10–16 or 1 in 2.48 quadrillion.Table 3 Coancestry (θ) and inbreeding (f) coefficients for Shorea leprosula at each hierarchical level.Full size table More

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.This is a summary of: Qiao, L. et al. Soil quality both increases crop production and improves resilience to climate change. Nat. Clim. Change https://doi.org/10.1038/s41558-022-01376-8 (2022). More

Brondizio, E. S., Settele, J., Díaz, S. & Ngo, H. T. (eds.) Global Assessment Report on Biodiversity and Ecosystem Services of the Intergovernmental Science–Policy Platform on Biodiversity and Ecosystem Services (IPBES Secretariat, 2019).Weiskopf, S. R. et al. Climate change effects on biodiversity, ecosystems, ecosystem services, and natural resource management in the United States. Sci. Total Environ. 733, 137782. https://doi.org/10.1016/j.scitotenv.2020.137782 (2020).ADS 
CAS 
Article 
PubMed 

Google Scholar 
Turner, J. & Marshall, G. J. Climate Change in the Polar Regions (Cambridge University Press, 2011).Book 

Google Scholar 
Meredith, M. et al. Polar Regions. Chapter 3, IPCC Special Report on the Ocean and Cryosphere in a Changing Climate. https://www.ipcc.ch/srocc/chapter/chapter-3-2/ (2019).Rignot, E. et al. Four decades of Antarctic Ice Sheet mass balance from 1979–2017. Proc. Natl. Acad. Sci. USA 116, 1095–1103. https://doi.org/10.1073/pnas.1812883116 (2019).ADS 
CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Siegert, M. et al. The Antarctic Peninsula under a 1.5°C global warming scenario. Front. Environ. Sci. 7, 102. https://doi.org/10.3389/fenvs.2019.00102 (2019).Article 

Google Scholar 
Iz, H. B. Is the global sea surface temperature rise accelerating?. Geod. Geodyn. 9, 432–438. https://doi.org/10.1016/j.geog.2018.04.002 (2018).Article 

Google Scholar 
Qiu, Z. et al. Future climate change is predicted to affect the microbiome and condition of habitat-forming kelp. Proc. R. Soc. B. 286, 20181887. https://doi.org/10.1098/rspb.2018.1887 (2019).Article 
PubMed 
PubMed Central 

Google Scholar 
Burge, C. A., Kim, C. J., Lyles, J. M. & Harvell, C. D. Special issue Oceans and Humans Health: The ecology of marine opportunists. Microb. Ecol. 65, 869–879. https://doi.org/10.1007/s00248-013-0190-7 (2013).Article 
PubMed 

Google Scholar 
Cavicchioli, R. et al. Scientists’ warning to humanity: Microorganisms and climate change. Nat. Rev. Microbiol. 17, 569–586. https://doi.org/10.1038/s41579-019-0222-5 (2019).CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Harvell, C. D. et al. Emerging marine diseases–climate links and anthropogenic factors. Science 285, 1505–1510. https://doi.org/10.1126/science.285.5433.1505 (1999).CAS 
Article 
PubMed 

Google Scholar 
Egan, S. & Gardiner, M. Microbial dysbiosis: Rethinking disease in marine ecosystems. Front. Microbiol. 7, 991. https://doi.org/10.3389/fmicb.2016.00991 (2016).Article 
PubMed 
PubMed Central 

Google Scholar 
Wilkins, L. G. E. et al. Host-associated microbiomes drive structure and function of marine ecosystems. PLoS Biol. 17, e3000533. https://doi.org/10.1371/journal.pbio.3000533 (2019).CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Seuront, L., Nicastro, K. R., Zardi, G. I. & Goberville, E. Decreased thermal tolerance under recurrent heat stress conditions explains summer mass mortality of the blue mussel Mytilus edulis. Sci. Rep. 9, 17498. https://doi.org/10.1038/s41598-019-53580-w (2019).ADS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Tsuchiya, M. Mass mortality in a population of the mussel Mytilus edulis L. caused by high temperature on rocky shores. J. Exp. Mar. Biol. Ecol. 66, 101–111. https://doi.org/10.1016/0022-0981(83)90032-1 (1983).Article 

Google Scholar 
Malham, S. K. et al. Summer mortality of the Pacific oyster, Crassostrea gigas, in the Irish Sea: The influence of temperature and nutrients on health and survival. Aquaculture 287, 128–138. https://doi.org/10.1016/j.aquaculture.2008.10.006 (2009).CAS 
Article 

Google Scholar 
Beyer, J. et al. Blue mussels (Mytilus edulis spp.) as sentinel organisms in coastal pollution monitoring: A review. Mar. Environ. Res. 130, 338–365. https://doi.org/10.1016/j.marenvres.2017.07.024 (2017).ADS 
CAS 
Article 
PubMed 

Google Scholar 
Ladeiro, M. P. et al. Mussel as a tool to define continental watershed quality. In Organismal and Molecular Malacology (ed Ray, S.), IntechOpen. https://doi.org/10.5772/67995 (2017).Bonacci, S. et al. Esterase activities in the bivalve mollusc Adamussium colbecki as a biomarker for pollution monitoring in the Antarctic marine environment. Mar. Pollut. Bull. 49, 445–455. https://doi.org/10.1016/j.marpolbul.2004.02.033 (2004).CAS 
Article 
PubMed 

Google Scholar 
Storhaug, E. et al. Seasonal and spatial variations in biomarker baseline levels within Arctic populations of mussels (Mytilus spp.). Sci. Total Environ. 656, 921–936. https://doi.org/10.1016/j.scitotenv.2018.11.397 (2019).ADS 
CAS 
Article 
PubMed 

Google Scholar 
Caza, F. et al. Liquid biopsies for omics-based analysis in sentinel mussels. PLoS ONE 14, e0223525. https://doi.org/10.1371/journal.pone.0225359 (2019).CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Ignatiadis, M., Sledge, G. W. & Jeffrey, S. S. Liquid biopsy enters the clinic – implementation issues and future challenges. Nat. Rev. Clin. Oncol. 18, 297–312. https://doi.org/10.1038/s41571-020-00457-x (2021).Article 
PubMed 

Google Scholar 
Kowarsky, M. et al. Numerous uncharacterized and highly divergent microbes which colonize humans are revealed by circulating cell-free DNA. Proc. Natl. Acad. Sci. USA 114, 9623–9628. https://doi.org/10.1073/pnas.1707009114 (2017).CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Chen, H. et al. Circulating microbiome DNA: An emerging paradigm for cancer liquid biopsy. Cancer Lett. 521, 82–87. https://doi.org/10.1016/j.canlet.2021.08.036 (2021).CAS 
Article 
PubMed 

Google Scholar 
Lokmer, A. et al. Spatial and temporal dynamics of Pacific oyster hemolymph microbiota across multiple scales. Front. Microbiol. 7, 1367. https://doi.org/10.3389/fmicb.2016.01367 (2016).Article 
PubMed 
PubMed Central 

Google Scholar 
Lokmer, A. & Wegner, M. K. Hemolymph microbiome of Pacific oysters in response to temperature, temperature stress and infection. ISME J. 9, 670–682. https://doi.org/10.1038/ismej.2014.160 (2015).CAS 
Article 
PubMed 

Google Scholar 
Auguste, M. et al. Exposure to TiO2 nanoparticles induces shifts in the microbiota composition of Mytilus galloprovincialis hemolymph. Sci. Total Environ. 670, 129–137. https://doi.org/10.1016/j.scitotenv.2019.03.133 (2019).ADS 
CAS 
Article 
PubMed 

Google Scholar 
Vezzulli, L. et al. Climate influence on Vibrio and associated human diseases during the past half-century in the coastal North Atlantic. Proc. Natl. Acad. Sci. USA 113, E5062–E5071. https://doi.org/10.1073/pnas.1609157113 (2016).CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Musella, M. et al. Tissue-scale microbiota of the Mediterranean mussel (Mytilus galloprovincialis) and its relationship with the environment. Sci. Total Environ. 717, 137209. https://doi.org/10.1016/j.scitotenv.2020.137209 (2020).ADS 
CAS 
Article 
PubMed 

Google Scholar 
Féral, J.-P. et al. PROTEKER: Implementation of a submarine observatory at the Kerguelen islands (Southern Ocean). Underw. Technol. 34, 3–10. https://doi.org/10.3723/ut.34.003 (2016).Article 

Google Scholar 
Spain, E. A. et al. Shallow seafloor gas emissions near Heard and McDonald Islands on the Kerguelen Plateau, southern Indian Ocean. Earth Space Sci. 7, e2019EA000695. https://doi.org/10.1029/2019EA000695 (2020).ADS 
Article 

Google Scholar 
Cao, S. et al. Structure and function of the Arctic and Antarctic marine microbiota as revealed by metagenomics. Microbiome. 8, 47. https://doi.org/10.1186/s40168-020-00826-9 (2020).CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Wang, L.-Y. et al. Comparison of bacterial community in aqueous and oil phases of water-flooded petroleum reservoirs using pyrosequencing and clone library approaches. Appl. Microbiol. Biotechnol. 98, 4209–4221. https://doi.org/10.1007/s00253-013-5472-y (2014).CAS 
Article 
PubMed 

Google Scholar 
Gutierrez, T., Berry, D., Teske, A. & Aitken, M. D. Enrichment of Fusobacteria in sea surface oil slicks from the Deepwater Horizon oil spill. Microorganisms. 4, 24. https://doi.org/10.3390/microorganisms4030024 (2016).CAS 
Article 
PubMed Central 

Google Scholar 
Michelou, V. K., Caporaso, J. G., Knight, R. & Palumbi, S. R. The ecology of microbial communities associated with Macrocystis pyrifera. PLoS ONE 8, e67480. https://doi.org/10.1371/annotation/48e29578-a073-42e7-bca4-2f96a5998374 (2013).ADS 
CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Florez, J. Z. et al. Structure of the epiphytic bacterial communities of Macrocystis pyrifera in localities with contrasting nitrogen concentrations and temperature. Algal Res. 44, 101706. https://doi.org/10.1016/j.algal.2019.101706 (2019).Article 

Google Scholar 
Minich, J. J. et al. Elevated temperature drives kelp microbiome dysbiosis, while elevated carbon dioxide induces water microbiome disruption. PLoS ONE 13, e0192772. https://doi.org/10.1371/journal.pone.0192772 (2018).CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Lin, J. D., Lemay, M. A. & Parfrey, L. W. Diverse bacteria utilize alginate within the microbiome of the giant kelp Macrocystis pyrifera. Front. Microbiol. 9, 1914. https://doi.org/10.3389/fmicb.2018.01914 (2018).Article 
PubMed 
PubMed Central 

Google Scholar 
Pierce, M. L. & Ward, J. E. Microbial ecology of the Bivalvia, with an emphasis on the family Ostreidae. J. Shellfish Res. 37, 793–806. https://doi.org/10.2983/035.037.0410 (2018).Article 

Google Scholar 
Pierce, M. L. & Ward, J. E. Gut Microbiomes of the Eastern Oyster (Crassostrea virginica) and the Blue Mussel (Mytilus edulis): Temporal variation and the influence of marine aggregate-associated microbial communities. mSphere. 4, e00730-19. https://doi.org/10.1128/mSphere.00730-19 (2019).Article 
PubMed 
PubMed Central 

Google Scholar 
Delille, D. & Gleizon, F. Distribution of enteric bacteria in Antarctic seawater surrounding the Port-aux-Francais permanent station (Kerguelen Island). Mar. Pollut. Bull. 46, 1179–1183. https://doi.org/10.1016/S0025-326X(03)00164-4 (2003).CAS 
Article 
PubMed 

Google Scholar 
Nguyen, T. V. & Alfaro, A. C. Metabolomics investigation of summer mortality in New Zealand Greenshell mussels (Perna canaliculus). Fish Shellfish Immunol. 106, 783–791. https://doi.org/10.1016/j.fsi.2020.08.022 (2020).CAS 
Article 
PubMed 

Google Scholar 
Vezzulli, L. et al. Comparative 16SrDNA gene-based microbiota profiles of the Pacific oyster (Crassostrea gigas) and the Mediterranean Mussel (Mytilus galloprovincialis) from a shellfish farm (Ligurian Sea, Italy). Microb. Ecol. 75, 495–504. https://doi.org/10.1007/s00248-017-1051-6 (2018).CAS 
Article 
PubMed 

Google Scholar 
Romalde, J. L., Diéguez, A. L., Lasa, A. & Balboa, S. New Vibrio species associated to molluscan microbiota: A review. Front. Microbiol. 4, 413. https://doi.org/10.3389/fmicb.2013.00413 (2014).Article 
PubMed 
PubMed Central 

Google Scholar 
Narayan, N. R. et al. Piphillin predicts metagenomic composition and dynamics from DADA2-corrected 16S rDNA sequences. BMC Genom. 21, 56. https://doi.org/10.1186/s12864-019-6427-1 (2020).CAS 
Article 

Google Scholar 
Peng, W. et al. Integrated 16S rRNA sequencing, metagenomics, and metabolomics to characterize gut microbial composition, function, and fecal metabolic phenotype in non-obese type 2 diabetic Goto-Kakizaki rats. Front. Microbiol. 10, 3141. https://doi.org/10.3389/fmicb.2019.03141 (2020).Article 
PubMed 
PubMed Central 

Google Scholar 
Koner, S. et al. Assessment of carbon substrate catabolism pattern and functional metabolic pathway for microbiota of limestone caves. Microorganisms 9, 1789. https://doi.org/10.21203/rs.3.rs-549787/v1 (2021).CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Li, Y. F. et al. Temperature elevation and Vibrio cyclitrophicus infection reduce the diversity of haemolymph microbiome of the mussel Mytilus coruscus. Sci. Rep. 9, 16391. https://doi.org/10.1038/s41598-019-52752-y (2019).ADS 
CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Scanes, E. et al. Climate change alters the haemolymph microbiome of oysters. Mar. Pollut. Bull. 164, 111991. https://doi.org/10.1016/j.marpolbul.2021.111991 (2021).CAS 
Article 
PubMed 

Google Scholar 
Hylander, B. L. & Repasky, E. A. Temperature as a modulator of the gut microbiome: What are the implications and opportunities for thermal medicine?. Int. J. Hyperth. 36, 83–89. https://doi.org/10.1080/02656736.2019.1647356 (2019).CAS 
Article 

Google Scholar 
Lo Giudice, A. et al. Marine bacterioplankton diversity and community composition in an antarctic coastal environment. Microb. Ecol. 63, 210–223. https://doi.org/10.1007/s00248-011-9904-x (2012).Article 
PubMed 

Google Scholar 
Yumoto, I. et al. Temperature and nutrient availability control growth rate and fatty acid composition of facultatively psychrophilic Cobetia marina strain L-2. Arch. Microbiol. 181, 345–351. https://doi.org/10.1007/s00203-004-0662-8 (2004).CAS 
Article 
PubMed 

Google Scholar 
Weingarten, E. A., Atkinson, C. L. & Jackson, C. R. The gut microbiome of freshwater Unionidae mussels is determined by host species and is selectively retained from filtered seston. PLoS ONE 14, e0224796. https://doi.org/10.1371/journal.pone.0224796 (2019).CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Rosa, M., Ward, J. E. & Shumway, S. E. Selective capture and ingestion of particles by suspension-feeding bivalve molluscs: A review. J. Shellfish Res. 37, 727–746. https://doi.org/10.2983/035.037.0405 (2018).Article 

Google Scholar 
Griffiths, C. L. & King, J. A. Some relationships between size, food availability and energy balance in the ribbed mussel Aulacomya ater. Mar. Biol. 51, 141–149. https://doi.org/10.1007/BF00555193 (1979).Article 

Google Scholar 
Riisgård, H. U. Filtration rate and growth in the blue mussel, Mytilus edulis Linneaus, 1758: Dependence on algal concentration. J. Shellfish Res. 10, 29–36 (1991).
Google Scholar 
Sonier, R. et al. Picophytoplankton contribution to Mytilus edulis growth in an intensive culture environment. Mar. Biol. 163, 73. https://doi.org/10.1007/s00227-016-2845-7 (2016).Article 

Google Scholar 
Jacobs, P., Troost, K., Riegman, R. & Van der Meer, J. Length-and weight-dependent clearance rates of juvenile mussels (Mytilus edulis) on various planktonic prey items. Helgol. Mar. Res. 69, 101–112. https://doi.org/10.1007/s10152-014-0419-y (2015).ADS 
Article 

Google Scholar 
Ward, J. E. & Shumway, S. E. Separating the grain from the chaff: Particle selection in suspension- and deposit-feeding bivalves. J. Exp. Mar. 300, 83–130. https://doi.org/10.1016/j.jembe.2004.03.002 (2004).Article 

Google Scholar 
Waite, A. M., Safi, K. A., Hall, J. A. & Nodder, S. D. Mass sedimentation of picoplankton embedded in organic aggregates. Limnol. Oceanogr. 45, 87–97. https://doi.org/10.4319/lo.2000.45.1.0087 (2000).ADS 
Article 

Google Scholar 
Ward, J. E. & Kach, D. J. Marine aggregates facilitate ingestion of nanoparticles by suspension-feeding bivalves. Mar. Environ. Res. 68, 137–142. https://doi.org/10.1016/j.marenvres.2009.05.002 (2009).CAS 
Article 
PubMed 

Google Scholar 
Ward, J. E. Biodynamics of suspension-feeding in adult bivalve molluscs: Particle capture, processing, and fate. Invertebr. Biol. 115, 218–231. https://doi.org/10.2307/3226932 (1996).Article 

Google Scholar 
Rosa, M. et al. Physicochemical surface properties of microalgae and their combined effects on particle selection by suspension-feeding bivalve molluscs. J. Exp. Mar. 486, 59–68. https://doi.org/10.1016/j.jembe.2016.09.007 (2017).CAS 
Article 

Google Scholar 
Allam, B. & Espinosa, E. P. Bivalve immunity and response to infections: Are we looking at the right place?. Fish Shellfish Immunol. 53, 4–12. https://doi.org/10.1016/j.fsi.2016.03.037 (2016).CAS 
Article 
PubMed 

Google Scholar 
Barr, J. J. et al. Bacteriophage adhering to mucus provide a non-host-derived immunity. Proc. Natl. Acad. Sci. USA 110, 10771–10776. https://doi.org/10.1073/pnas.1305923110 (2013).ADS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Allam, B. & Espinosa, E. P. Mucosal immunity in mollusks. In Mucosal Health in Aquaculture (eds Beck, B. H. & Peatman, E.) 325–370 (Academic Press, 2015).Chapter 

Google Scholar 
Huang, J. et al. Hemocytes in the extrapallial space of Pinctada fucata are involved in immunity and biomineralization. Sci. Rep. 8, 4657. https://doi.org/10.1038/s41598-018-22961-y (2018).ADS 
CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Kim, H. J. et al. Isolation and characterization of two bacteriophages and their preventive effects against pathogenic Vibrio coralliilyticus causing mortality of Pacific oyster (Crassostrea gigas) larvae. Microorganisms. 8, 926. https://doi.org/10.3390/microorganisms8060926 (2020).CAS 
Article 
PubMed Central 

Google Scholar 
Ihara, H. et al. Sulfur-oxidizing bacteria mediate microbial community succession and element cycling in launched marine sediment. Front. Microbiol. 8, 152. https://doi.org/10.3389/fmicb.2017.00152 (2017).ADS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Jørgensen, B. B. & Nelson, D. C. Sulfide oxidation in marine sediments: Geochemistry meets microbiology. Geol. S. Am. S. 379, 63–81. https://doi.org/10.1130/0-8137-2379-5.63 (2004).Article 

Google Scholar 
Zhou, M. et al. Surface currents and upwelling in Kerguelen Plateau regions. Biogeosci. Discuss. 11, 6845–6876. https://doi.org/10.5194/bgd-11-6845-2014 (2014).ADS 
Article 

Google Scholar 
Gille, S. T., Carranza, M. M., Cambra, R. & Morrow, R. Wind-induced upwelling in the Kerguelen Plateau region. Biogeosciences 11, 6389–6400. https://doi.org/10.5194/bg-11-6389-2014 (2014).ADS 
Article 

Google Scholar 
Park, Y. H., Roquet, F., Durand, I. & Fuda, J. L. Large-scale circulation over and around the Northern Kerguelen Plateau. Deep Sea Res. II(55), 566–581. https://doi.org/10.1016/j.dsr2.2007.12.030 (2008).ADS 
Article 

Google Scholar 
Renac, C. et al. Hydrothermal fluid interaction in basaltic lava units, Kerguelen Archipelago (SW Indian Ocean). Eur. J. 22, 215–234. https://doi.org/10.1127/0935-1221/2009/0022-1993 (2010).CAS 
Article 

Google Scholar 
Vancanneyt, M. et al. Sphingomonas alaskensis sp. nov., a dominant bacterium from a marine oligotrophic environment. Int. J. Syst. Evol. 51, 73–79. https://doi.org/10.1099/00207713-51-1-73 (2001).CAS 
Article 

Google Scholar 
Helmuth, B. S. & Hofmann, G. E. Microhabitats, thermal heterogeneity, and patterns of physiological stress in the rocky intertidal zone. Biol. Bull. 201, 374–384. https://doi.org/10.2307/1543615 (2001).CAS 
Article 
PubMed 

Google Scholar 
Testut, L., Wöppelmann, G., Simon, B. & Téchiné, P. The sea level at Port-aux-Français, Kerguelen Island, from 1949 to the present. Ocean Dyn. 56, 464–472. https://doi.org/10.1007/s10236-005-0056-8 (2006).ADS 
Article 

Google Scholar 
Pohl, B. et al. Recent climate variability around the Kerguelen Islands (Southern Ocean) seen through weather regimes. J. Appl. Meteorol. Climatol. 60, 711–731. https://doi.org/10.1175/JAMC-D-20-0255.1 (2021).ADS 
Article 

Google Scholar 
PROTEKER. Ilôt Channer (Passe Royale)—Sea water temperature at 5 and 13 m depth (T°C) daily average 2014–2019. https://www.proteker.net/swt-ilot-channer-passe-royale/ (2021).Caza, F. et al. Comparative analysis of hemocyte properties from Mytilus edulis desolationis and Aulacomya ater in the Kerguelen Islands. Mar. Environ. Res. 110, 174–182. https://doi.org/10.1016/j.marenvres.2015.09.003 (2015).CAS 
Article 
PubMed 

Google Scholar 
Caza, F., Cledon, M. & St-Pierre, Y. Biomonitoring climate change and pollution in marine ecosystems: A review on Aulacomya ater. J. Mar. Biol. 2016, 7183813. https://doi.org/10.1155/2016/7183813 (2016).Article 

Google Scholar 
Rey-Campos, M. et al. High individual variability in the transcriptomic response of Mediterranean mussels to Vibrio reveals the involvement of myticins in tissue injury. Sci. Rep. 9, 3569. https://doi.org/10.1038/s41598-019-39870-3 (2019).ADS 
CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Caza, F. et al. Hemocytes released in seawater act as Trojan horses for spreading of bacterial infections in mussels. Sci. Rep. 10, 19696. https://doi.org/10.1038/s41598-020-76677-z (2020).ADS 
CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Yao, C. L. & Somero, G. N. Thermal stress and cellular signaling processes in hemocytes of native (Mytilus californianus) and invasive (M. galloprovincialis) mussels: Cell cycle regulation and DNA repair. Comp. Biochem. Physiol. 165, 159–168. https://doi.org/10.1016/j.cbpa.2013.02.024 (2013).CAS 
Article 

Google Scholar 
Lockwood, B. L., Sanders, J. G. & Somero, G. N. Transcriptomic responses to heat stress in invasive and native blue mussels (genus Mytilus): Molecular correlates of invasive success. J. Exp. Biol. 213, 3548–3558. https://doi.org/10.1242/jeb.046094 (2010).CAS 
Article 
PubMed 

Google Scholar 
Klindworth, A. et al. Evaluation of general 16S ribosomal RNA gene PCR primers for classical and next-generation sequencing-based diversity studies. Nucleic Acids Res. 41, e1. https://doi.org/10.1093/nar/gks808 (2013).CAS 
Article 
PubMed 

Google Scholar 
Callahan, B. J. et al. DADA2: High-resolution sample inference from Illumina amplicon data. Nat. Methods. 13, 581–583. https://doi.org/10.1038/nmeth.3869 (2016).CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
R Core Team. R: A Language and Environment for Statistical Computing (R Foundation for Statistical Computing, 2021).
Google Scholar 
McMurdie, P. J. & Holmes, S. phyloseq: An R package for reproducible interactive analysis and graphics of microbiome census data. PLoS ONE 8, e61217. https://doi.org/10.1371/journal.pone.0061217 (2013).ADS 
CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Oksanen, J. & Blanchet, F. G. Vegan: Community Ecology Package. 2. 3-0 (2015).Ssekagiri, A., Sloan, W. & Ijaz, U. Z. microbiomeSeq: an R package for analysis of microbial communities in an environmental context, In ISCB Africa ASBCB Conference (Kumasi, Ghana, 2017).Cao, Y. Microbiome marker: Microbiome Biomarker Analysis Toolkit. R package version 0.99.0 (2020). https://github.com/yiluheihei/microbiomeMarker. Accessed March 2022.Kanehisa, M. et al. KEGG for integration and interpretation of large-scale molecular data sets. Nucleic Acids Res. 40, D109–D114. https://doi.org/10.1093/nar/gkr988 (2012).CAS 
Article 
PubMed 

Google Scholar 
Iwai, S. et al. Piphillin: Improved prediction of metagenomic content by direct inference from human microbiomes. PLoS ONE 11, e0166104. https://doi.org/10.1371/journal.pone.0166104 (2016).CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Dhariwal, A. et al. MicrobiomeAnalyst: A web-based tool for comprehensive statistical, visual and meta-analysis of microbiome data. Nucleic Acids Res. 45, W180–W188. https://doi.org/10.1093/nar/gkx295 (2017).CAS 
Article 
PubMed 
PubMed Central 

Google Scholar  More

Chloroplast and repetitive nuclear DNA enrichment in the sedaDNA extractsTo the best of our knowledge, we generated the first large-scale target enriched dataset using sedaDNA extracted from sediments of multiple lakes. Sequencing of two datasets produced 325.5 million (M) quality-filtered paired-end DNA sequences. The first target enriched dataset, targeting both the chloroplast and a set of nuclear genes of Larix on 64 sedaDNA extracts and 19 negative controls from seven lake sediment records resulted in 324 M quality-filtered paired-end sequences. The second target enriched dataset, targeting only the set of nuclear genes of Larix on four samples and two negative controls from an additional lake (Lake CH12) resulted in 1.5 M sequences. Quality-filtering of an additional published target enriched dataset29, targeting the Larix chloroplast genome on the same CH12 samples as applied for the second dataset, added another 54 M sequences.For the chloroplast enrichment, 390 thousand (K) sequences (1%) were classified as Larix at the genus or species level. The average coverage of bait regions was 19% at a mean sequence depth of 0.8. Sequencing of 19 library and extraction blank (negative control) samples resulted in 597 K paired-end sequences, of which 58% quality-filtered and deduplicated sequences remained. Of these, 38% were classified, with 0.03% of them (463 sequences) corresponding to the genus Larix. Negative controls from library preparation resulted in no to very few (0 to 5) sequences mapping to the Larix chloroplast reference genome. Negative controls from DNA extractions, which were in several cases pooled to one library, showed a low number of sequences mapped to Larix (0 to 94 sequences, except 237 sequences in one case). Excluding all sequences in negative controls from the sample analysis had no impact on the patterns resulting from the analysis of sample data. Detailed results and evaluation of negative controls are included in the Supplementary Information (Fig. S5) and Supplementary Data 1 and 2. Samples of all lake records with sufficient sequence coverage showed damage patterns typical of ancient DNA (see Supplementary Data 3).These results are comparable to the results obtained by Schulte et al.29, where 36% of quality-filtered sequences were classified as Viridiplantae with 9% assigned to Larix. In contrast to29, we raised the confidence threshold of taxonomic classification (a parameter defining the number of k-mers needed to produce a match against a taxon in the database), which drastically reduced the number of classified sequences, but increased the confidence in the analysis36.To analyze the enrichment obtained by the nuclear gene bait set, taxonomic classification was repeated using a plant genome database including available Pinaceae genomes. The classification resulted in 716 K sequences assigned to Larix, increasing the previous results by 325 K sequences. However, almost no sequences were mapped against the targeting baits (a maximum of five sequences for some samples). A closer inspection of unmapped sequences assigned to Larix revealed a high content of repetitive DNAs. More specifically, taxonomically classified Larix sequences could be assembled to EulaSat1, the most abundant satellite repeat in the nuclear genome of Larix32,37. This short repeat with a 173 bp long motif is arranged in large arrays of tandemly repeated motifs and is exclusively present in larches32. Analysis of modern L. sibirica, and L. gmelinii (western and eastern range) genomes reveals that EulaSat1 occurs in all species, contributing to 0.62% (L. sibirica), 2.52% (western range L. gmelinii), and 2.39% (eastern range L. gmelinii), of the genomes, respectively (Fig. S2). A comparison of the sequence proportions mapping to the repeat motif in the different datasets of Lake CH12 showed a specific enrichment of the repeat motif by the nuclear gene hybridization probe set (Fig. S3).In total, 17 K sequences mapped to the repeat motif of EulaSat1. The abundance of all sequences mapped per sample is in agreement with the abundance of sequences mapped to the chloroplast genome, confirming the general history of forest development (Fig. 2). Analysis of the nucleotide frequencies in the repeat motif showed a high constancy over all samples (Fig. S4). This suggests high conservation of the EulaSat1 motif in Siberian larches over time and space. Although satellite repeats are reported to have a high sequence turnover, for larches it has been shown that repeat profiles between two geographically well-separated species—the European larch (L. decidua) and the Japanese larch (L. kaempferi)—are very similar32. The main satellite in all larches, EulaSat1, is believed to have greatly multiplied after the split of Larix from Pseudotsuga32. Given the ongoing hybridization between the three Siberian larch species, it is not surprising to find a consistent pattern of nucleotide frequencies in all samples.Fig. 2: Comparison of target enrichment with available DNA metabarcoding and pollen datasets.From left to right: Larix-classified sequence counts mapping to (1) the Larix chloroplast and (2) the EulaSat1 satellite repeat motif, (3) percentage of Larix counts in metabarcoding data, (4) percentage of Larix pollen in pollen assemblages. All data from this study, except metabarcoding data from lakes CH1213 and Bolshoye Shchuchye55 and all pollen data except for several samples of Lake Kyutyunda which were produced in this study56,57,71. Pollen data of Lake Lama and the Holocene part of Lake Kyutyunda are based on parallel sediment cores PG1111 and PG2022, respectively. No available data are marked with crosses, asterisk marks a single Larix pollen grain found in the Bolshoye Shchuchye sediments.Full size imageOff-target sequences in target enriched datasets have already been demonstrated to be useful for the analysis of high-copy DNA such as ribosomal DNA or plastomes34,38,39. A recent study on five modern sedges showed that target enriched sequencing data originally targeting a set of gene exons can also be used to study the repetitive sequence fraction and even infer phylogenetic relationships based on repetitive sequence abundance35. Another study showed that also sequence similarities between homologous repeat motifs can be used to reconstruct phylogenetic relationships among closely related taxa40,41. In the case of Larix satellite EuLaSat1 in our study, no change in nucleotide frequencies, neither related to locations nor in time, could be detected. However, our results show that the off-target fraction in target enriched sedaDNA datasets can hold valuable information and that repeat motifs in more diverse taxon groups could even be a target for enrichment. Specifically enriching for repeat motifs in sedaDNA extracts could enable the study of satellite repeat evolution as well as giving additional information on species abundance and phylogeography.In the two target enriched datasets, sequences taxonomically classified to the genus Larix and mapping to the chloroplast and to the repeat sequence, respectively, show similar patterns of abundance (see Fig. 2). Compared with published metabarcoding and pollen data from the same locations, the Larix abundance patterns can be globally reproduced, underpinning the notion that sequence abundances in target enriched data can be used as good estimates of plant abundances. For older parts of the lake records, target enriched data show Larix where metabarcoding data were unable to detect a clear signal (see Fig. 2, lakes Billyakh, Bolshoye Shchuchye, Kyutyunda, and Lama). This shows that target enrichment is superior to metabarcoding when analyzing one taxonomic group in-depth, as it is less prone to errors by DNA degradation, which can impede primer binding if the molecule becomes too short. Also, independent of age, rare taxa mostly need multiple PCR replicates to be detected by metabarcoding42,43. Target enrichment, however, is more sensitive in identifying one focal taxon group, as the total target length can be much larger (e.g., a complete organellar genome) than for metabarcoding, and the DNA damage patterns are put to use to authenticate ancient DNA. Also, it is limited by molecule length only by the applied threshold in the bioinformatic analysis, for which we used 30 base pairs (bp) as opposed to a minimum of 85 bp molecule length for the Larix metabarcoding marker (for the plant-specific trnL g/h marker44). Similarly, compared to traditional pollen analysis, target enrichment is more accurate at tracing a specific target group, as it is not dependent on pollen productivity. Especially in the case of Larix, pollen productivity is low and preservation poor, resulting in rare findings of its pollen in the sediments22,45. This could explain why for Lake Bolshoye Shchuchye, only a single Larix pollen grain was retrieved throughout the core, whereas target enrichment and metabarcoding show a strong signal in the Holocene sediments (last ~12 ka BP). Target-enriched data also records signals in MIS 2 sediments, however, sequence counts are extremely low, and as it is the only record, where both of the other proxies fail to report a signal, it should be interpreted with caution.A wider pre-glacial distribution of L. sibirica
Chloroplast genomes of L. gmelinii and L. sibirica differ at 157 positions, which can be used to differentiate species in target enriched sedaDNA29. Here, we applied this approach to lake sediment records, which are distributed across Siberia (Fig. 1) and have time ranges back to MIS3, and thereby were able to track species composition in space and time for wide parts of the species ranges.In lakes Billyakh and Kyutyunda, ca. 1500 km east of L. sibirica current range (Fig. 1), we found evidence for a wider distribution of L. sibirica around 32 and 34 ka BP in MIS3 (Fig. 3). Billyakh is situated in the western part of the Verkhoyansk Mountains, and Kyutyunda on the Central Siberian Plateau. Both lakes have low counts of Larix DNA sequences in their oldest samples dated to 51 ka BP (Billyakh) and 38 ka BP (Kyutyunda) with variants of L. gmelinii, but there is a sudden rise in variants attributed to L. sibirica at 34 ka BP (Billyakh) and 32 ka BP (Kyutyunda), which persists in the following samples, but strongly decreases in younger samples (Fig. 3). The rise in the L. sibirica DNA sequence variants coincides with a peak in sequence counts for Lake Kyutyunda. These signals suggest a rapid invasion of L. sibirica into the ranges of L. gmelinii in climatically favorable times and a local depletion or extinction of L. sibirica during the following harsher climates. Lake Billyakh pollen data suggest a moister and warmer climate around 50–30 ka BP than in the latter part of the Last Glacial associated with the MIS3 Interstadial in Siberia46.Fig. 3: Percentage and sequence counts at variable positions along Larix chloroplast genome assigned to species.Left: Alignment of Larix-classified DNA sequences against the chloroplast genome at the 157 variable positions between the species. For each position, the percentage of sequences assigned to a single species is displayed. Each row represents one sample named according to the calibrated age before present. Gray background indicates no coverage at the respective position. Right: Total number of sequences assigned to each of the species per sample.Full size imageStrong support for a wider pre-glacial distribution of L. sibirica comes from genetic analyses which show that it is genetically close to L. olgensis, today occurring on the Korean Peninsula and adjacent areas of China and Russia27,47. It is assumed that the L. sibirica-L. olgensis complex used to share a common range, which was disrupted and displaced when the better cold-adapted L. gmelinii expanded south and southwest during the more continental climatic conditions of the Pleistocene47,48. Furthermore, modern and ancient genetic studies suggest that the L. sibirica zone was recently invaded by L. gmelinii from the east in the hybridization zone of the species, as the climate cooled after the mid-Holocene thermal maximum13,23. Today, pure stands of L. sibirica do not form a continuous habitat, but occur in netted islands5 and morphological features of L. sibirica can be found in populations of L. gmelinii located at least a hundred kilometers east of the closest L. sibirica populations49. Macrofossil findings of L. sibirica in Scandinavia dated to the early Holocene, point to the capability of rapid long-distance jump dispersal of this species50. Fossil L. sibirica cones dated to the end of the Pliocene and in the Pleistocene have also been found far east of its current range in several river valleys including Kolyma, Aldan, and Omolon, and even in the basin of the Sea of Okhotsk9. These indicate long-distance seed dispersal by rivers which may also have assisted in successful establishment since the active-layer depth is deeper close to rivers51,52. As mentioned earlier, L. sibirica is sensitive to permafrost and waterlogged soils. A warmer phase with a deeper thawed layer above the permafrost could have enabled L. sibirica to spread and establish in regions that today are part of the geographic range of L. gmelinii, as L. sibirica is reported to have higher growth rates than L. gmelinii13.
Larix gmelinii formed northern LGM refugia across SiberiaThe possible survival of Larix in high latitude glacial refugia during the LGM is still under discussion4,53 although more and more evidence is reported in favor of the existence of such refugia17,20,21. The question of which of the Larix species formed these populations has hitherto been unanswered, as both pollen and established metabarcoding markers are not able to distinguish between species in the genus Larix, and findings of fossilized cones identifiable to species are rare. By enriching sedaDNA extracts for chloroplast genome sequences, we are, to the best of our knowledge, for the first time, able to distinguish between L. sibirica and L. gmelinii in glacial refugial populations.From Lake Lama, located at the western margin of the Putorana Plateau (Taymyr Peninsula), we obtained a continuous record extending from 23 ka BP to today with varying sequence counts with minima around 18–17 ka BP and 13 ka BP. All samples prior to the Holocene show variations predominantly assigned to L. gmelinii (Fig. 3). Our results suggest a local survival of L. gmelinii in the vicinity of Lake Lama throughout the LGM, which is supported by low numbers of Larix pollen detected through this period. Both target enriched sequence data and pollen indicate an increase from ca. 11 ka BP54. Sparse Larix pollen in the bottom part of the record could be an indication of a possible refugial population (Fig. 2; ref. 54).In Bolshoye Shchuchye, the westernmost lake of the study, situated in the Polar Ural Mountains, all Pleistocene samples show similarly a dominance of L. gmelinii sequence variations (Fig. 3). However, sequence counts for some samples are extremely low and samples from 18 and 10 ka BP had so low counts of mapped DNA sequences that none of the variable positions between the species was covered. Although sequences mapped to the satellite repeat of Larix also showed a Pleistocene signal, this was not repeated in pollen or metabarcoding (Fig. 2) which instead indicates a treeless arctic-alpine flora for the late Pleistocene55,56. Especially for the sample of 20.4 ka, Larix sequence counts are extremely low and new investigations would be needed to confirm a local presence of Larix during the LGM.The record of Lake Billyakh situated in the western Verkhoyansk Mountain Range, likewise shows extremely low counts of sequences mapped to the reference for a range of samples with no sequences covering the studied variable sites (45, 42, and 15 ka BP, 11–56 sequences mapped to non-variable sites). However, the pollen record for the same core shows a quasi-continuous record of Larix with a gap only occurring during the early LGM46 (25–22 ka BP, Fig. 2). Considering the known short-distance dispersal ability and poor preservation of Larix pollen, this strongly supports the presumed existence of a local glacial refugium at Lake Billyakh during that time20. Our samples also show a low but steady presence of Larix throughout the rest of the record, thus making glacial survival probable. The sample closest to the LGM (24 ka BP) indicates a clear dominance of L. gmelinii type variations.The only exception to this general pattern is the record from Lake Kyutyunda, which is located on the Central Siberian Plateau west of the Verkhoyansk Mountain Range. In this record, LGM samples have extremely low counts but show variations assigned to L. sibirica and not to L. gmelinii as in the other lakes. In addition, the preceding sample dated to the MIS3 interstadial shows L. sibirica variation. A possible explanation is that relics of L. sibirica survived during the LGM, but were unable to spread after climate warming, possibly due to genetic depletion or later local extinction. The presence of reworked sediment material can also not be excluded, as suggested by reworked pollen in the record57.In conclusion, our data show almost exclusively L. gmelinii variation for samples covering the most severe LGM climate conditions. This is in agreement with the ecological characteristics describing the species as adapted to extreme cold. In contrast to L. sibirica, it can grow in dwarf forms and propagate clonally and potentially survive thousands of years of adverse climatic conditions58.Postglacial colonization history—differences among larch speciesOf great interest in the Larix history is not only the location and extent of possible high latitude glacial refugia but also if and to what extent these refugia contributed to the recolonization of Siberia after the LGM. Northern refugial populations could have functioned as kernels of postglacial population spread and recolonization, or spreading could have been driven by populations that survived in southern refugia. There are only a few studies on modern populations that report evidence for possible recolonization scenarios of Larix23,27,28. Here, we show that patterns differ between L. sibirica and L. gmelinii.In the western part of our study region, two lakes are situated in the current distribution range of L. sibirica (Figs. 1, 4): Lake Bolshoye Shchuchye in the Polar Ural Mountains and Lake Lama on the Taymyr Peninsula. Despite this, both lakes show L. gmelinii for all Pleistocene samples, and a strong signal of L. sibirica variants only in the Holocene samples, with ages of 5.1 ka BP in Lake Bolshoye Shchuchye and 9.7 ka BP in Lake Lama (Fig. 3). The peak in L. sibirica also coincides with a peak of sequence counts in the respective sample, with a Larix pollen peak in Lake Lama sediments54, and metabarcoding for Lake Bolshoye Shchuchye55. This points to a migration of L. sibirica in its current northern area of northern distribution in the course of climate warming during the early Holocene, whereas glacial refugial populations were consisting of L. gmelinii. Although the local survival of L. gmelinii around Lake Bolshoye Shchuchye remains uncertain due to extremely low sequence counts, it is clear that L. sibirica did not form a refugial population at this site.Fig. 4: Percentage of DNA sequences assigned to references displayed on the geographical locations of the lakes investigated.Samples in the same time frame are averaged. Lake names and current species ranges are annotated in the middle plots. Colors indicate current species distribution (adapted from Semerikov and Lascoux72). The base map is done with ggmap73, map tiles by Stamen Design, under CC BY 3.0. Data by OpenStreetMap, under ODbL.Full size imageA range-wide genetic study of L. sibirica analyzing chlorotypes and mitotypes of individuals23 found strong indications for rapid colonization of the West Siberian Plains from populations originating from the foothills of the Sayan Mountains in the south, close to the border of Mongolia, with only limited contribution from local populations. According to our results, the local populations could have been L. gmelinii populations, while the rapid invasion could have been L. sibirica.In the eastern range of the study region, in the current range of L. gmelinii, namely at lakes Emanda, Satagay, and Malaya Chabyda, genetic variations throughout the records are less pronounced. Of the three lake records, only that from Lake Emanda reaches back beyond the LGM, but with a sampling gap for the time of the LGM. Therefore, it remains uncertain whether populations survived the LGM locally, or whether they were invaded or replaced by populations coming from the south with Holocene warming. The restricted variations throughout the record, however, hint at stable populations, which is supported by scarce pollen findings (Fig. 2).Our data suggest that postglacial recolonization of L. sibirica was not started from high latitude glacial refugia, but from southern populations. In contrast, northern glacial populations of L. gmelinii could have potentially enhanced rapid dispersal after the LGM in their current area of distribution.Environment likely plays a more important role than historical factorsThe current boundaries of boreal Larix species arranged from west to east suggest a possible strong influence of the historical species distribution on the current distribution, whereas the gradient of increasing continental climate towards the east assumes a strong influence on the environment. By tracking species distribution in the past, spanning the time of the strongly adverse climate of the LGM, we can give hitherto unprecedented insights into species distribution history.Several lines of evidence suggest a strong influence of the environment on species distribution: (1) Signals for L. sibirica appeared in its current area of distribution as late as the Holocene warming, whereas cold Pleistocene samples are dominated by L. gmelinii type variation; (2) in lakes far east of its modern range, signals of variation typical for L. sibirica coincide with peaks in sequence counts (29 ka BP, Lake Billyakh; 32 ka BP Lake Kyutyunda), which point to more forested vegetation around the lakes and consequently a more favorable climate at that time; and (3) samples dated to the severely cold LGM are dominated by variations of the L. gmelinii type.This is in accordance with the different ecological characteristics described for the species. L. sibirica is sensitive to permafrost and only occurs outside of the zone of continuous permafrost5. In addition, L. sibirica achieves substantially higher growth rates and longer growth periods than L. gmelinii9,13 and can also produce more than twice as many seeds5. This potentially gives L. sibirica the ability to quickly react to climate change and outcompete the other species when the climate becomes more favorable.In contrast, L. gmelinii is adapted to extremely low soil and air temperatures and is able to grow on permafrost with very shallow thaw depths. It’s distribution almost completely coincides with continuous permafrost5, and even a restriction to permafrost areas is discussed as it does not grow well in field trials on warmer soils or where there is a small temperature gradient between air and soil9. Due to this ecology, L. gmelinii is more likely to survive in a high latitude refugium, even during the severe continental climate of the LGM, which was most probably connected to continuous permafrost of low active-layer depths.A study combining mitochondrial barcoding on sedaDNA and a modeling approach on Larix distribution in the Taymyr region around Lake CH12 concluded that the distributions of L. gmelinii and L. sibirica are most strongly influenced by stand density and thus by competition between the species, with L. gmelinii outcompeting L. sibirica at high stand densities13. As our study includes sediment cores reaching further back in time, we see a different trend. Instead of L. gmelinii, it was L. sibirica, which dominated samples with high sequence counts, suggesting high stand density and a more favorable climate. A possible explanation for the different outcomes is the use of different organelle genomes. Epp et al.13 used a marker representing the mitochondrial genome, which is known to introgress more rapidly and as a consequence might show a long past species history59,60.Our findings have potentially important implications for the projections of vegetation-climate feedback. A warming climate in conjunction with a greater permafrost thaw depth could enable the replacement of L. gmelinii by L. sibirica. In contrast to L. gmelinii, L. sibirica is not known to stabilize permafrost thus potentially further promoting permafrost thaw and with it the release of greenhouse gases, creating positive feedback on global warming11. On the other hand, the substantially higher growth rates of L. sibirica in comparison to L. gmelinii would increase carbon sequestration, thus mitigating global warming13. This shows the importance of understanding species-specific reactions to climate change, which can result in great shifts in distribution. Target enrichment applied on sedaDNA is able to reveal the impact of past climate change on populations and the increasing availability of modern reference genomes will further enhance its value of information. More

Game, E. T. et al. Pelagic protected areas: The missing dimension in ocean conservation. Trends Ecol. Evol. 24, 360–369 (2009).PubMed 
Article 

Google Scholar 
McCauley, D. J. et al. Marine defaunation: Animal loss in the global ocean. Science 347, 1255641–1255647 (2015).PubMed 
Article 
CAS 

Google Scholar 
Paleczny, M., Hammill, E., Karpouzi, V. & Pauly, D. Population trend of the world’s monitored seabirds, 1950–2010. PLoS ONE 10, e0129342 (2015).PubMed 
PubMed Central 
Article 
CAS 

Google Scholar 
Croxall, J. P. et al. Seabird conservation status and threats: A global assessment of priorities. Bird Conserv. Int. 22, 1–34 (2012).Article 

Google Scholar 
Dias, M. P. et al. Threats to seabirds: A global assessment. Biol. Conserv. 237, 525–537 (2019).Article 

Google Scholar 
Trathan, P. N. et al. Pollution, habitat loss, fishing, and climate change as critical threats to penguins. Conserv. Biol. 29, 31–41 (2014).PubMed 
Article 

Google Scholar 
Boersma, D. et al. Applying science to pressing conservation needs for penguins. Conserv. Biol. 34, 103–112 (2019).PubMed 
PubMed Central 
Article 

Google Scholar 
Ropert-Coudert, Y. et al. Happy feet in a hostile world? The future of penguins depends on proactive management of current and expected threats. Front. Mar. Sci. 6, 248 (2019).Article 

Google Scholar 
Maestro, M., Pérez-Cayeiro, M. L., Chica-Ruiz, J. A. & Reyes, H. Marine protected areas in the 21st century: Current situation and trends. Ocean Coast. Manag. 171, 28–36 (2019).Article 

Google Scholar 
Hays, G. C. et al. Key questions in marine megafauna movement ecology. Trends Ecol. Evol. 31, 463–475 (2016).PubMed 
Article 

Google Scholar 
Boyd, C. et al. Spatial scale and the conservation of threatened species. Conserv. Lett. 1, 37–43 (2008).Article 

Google Scholar 
Marra, P. P., Cohen, E. B., Loss, S. R., Rutter, J. E. & Tonra, C. M. A call for full annual cycle research in animal ecology. Biol. Lett. 11, 20150552 (2015).PubMed 
PubMed Central 
Article 
CAS 

Google Scholar 
Kürten, N. et al. High individual repeatability of the migratory behaviour of a long-distance migratory seabird. Mov. Ecol. 10, 5 (2022).PubMed 
PubMed Central 
Article 

Google Scholar 
Weimerskirch, H. et al. Lifetime foraging patterns of the wandering albatross: Life on the move!. J. Exp. Mar. Biol. Ecol. 450, 68–78 (2014).Article 

Google Scholar 
Trebilco, R., Gales, R., Baker, G. B., Terauds, A. & Sumner, M. D. At sea movement of Macquarie Island giant petrels: Relationships with marine protected areas and Regional Fisheries Management Organisations. Biol. Conserv. 141, 2942–2958 (2008).Article 

Google Scholar 
Clay, T. A. et al. A comprehensive large-scale assessment of fisheries bycatch risk to threatened seabird populations. J. Appl. Ecol. 56, 1882–1893 (2019).Article 

Google Scholar 
Meier, R. E. et al. Tracking, feather moult and stable isotopes reveal foraging behaviour of a critically endangered seabird during the non-breeding season. Divers. Distrib. 23, 130–145 (2017).Article 

Google Scholar 
Frankish, C. K., Phillips, R. A., Clay, T. A., Somveille, M. & Manica, A. Environmental drivers of movement in a threatened seabird: Insights from a mechanistic model and implications for conservation. Divers. Distrib. 26, 1315–1329 (2020).Article 

Google Scholar 
Ratcliffe, N. et al. Changes in prey fields increase the potential for spatial overlap between gentoo penguins and a krill fishery within a marine protected area. Divers. Distrib. 27, 552–563 (2021).Article 

Google Scholar 
Grémillet, D. et al. Persisting worldwide seabird-fishery competition despite seabird community decline. Curr. Biol. 28, 4009–4013 (2018).PubMed 
Article 
CAS 

Google Scholar 
Bogdanova, M. I. et al. Multi-colony tracking reveals spatio-temporal variation in carry-over effects between breeding success and winter. Mar. Ecol. Prog. Ser. 578, 167–181 (2017).Article 
ADS 

Google Scholar 
van Bemmelen, R. et al. Flexibility in otherwise consistent non-breeding movements of a long-distance migratory seabird, the long-tailed skua. Mar. Ecol. Prog. Ser. 578, 197–211 (2017).Article 
ADS 

Google Scholar 
Robinson, W. M. L., Butterworth, D. S. & Plagányi, É. E. Quantifying the projected impact of the South African sardine fishery on the Robben Island penguin colony. ICES J. Mar. Sci. 72, 1882–1883 (2015).Article 

Google Scholar 
Sherley, R. B. et al. Bottom-up effects of a no-take zone on endangered penguin demographics. Biol. Lett. 11, 20150237 (2015).PubMed 
PubMed Central 
Article 
CAS 

Google Scholar 
Studholme, K. R., Hipfner, J. M., Domalik, A. D., Ivrson, S. J. & Crossin, G. T. Year-round tracking reveals multiple migratory tactics in a sentinel North Pacific seabird, Cassin’s auklet. Mar. Ecol. Prog. Ser. 619, 169–185 (2019).Article 
ADS 

Google Scholar 
Salton, M., Saraux, C., Dann, P. & Chiaradia, A. Carry-over body mass effect from winter to breeding in a resident seabird, the little penguin. R. Soc. Open Sci. 2, 140390 (2015).PubMed 
PubMed Central 
Article 
ADS 

Google Scholar 
Barbraud, C. et al. Density dependence, prey accessibility and prey depletion by fisheries drive Peruvian seabird population dynamics. Ecography 41, 1092–1102 (2018).Article 

Google Scholar 
Grémillet, D. et al. Starving seabirds: Unprofitable foraging and its fitness consequences in Cape gannets competing with fisheries in the Benguela upwelling ecosystem. Mar. Biol. 163, 1–11 (2016).Article 

Google Scholar 
Cook, A. S. C. P., Dadam, D., Mitchell, I., Ross-Smith, V. H. & Robinson, R. A. Indicators of seabird reproductive performance demonstrate the impact of commercial fisheries on seabird populations in the North Sea. Ecol. Indic. 38, 1–11 (2014).Article 

Google Scholar 
Thiebot, J.-B. et al. Adjustment of pre-moult foraging strategies in Macaroni Penguins Eudyptes chrysolophus according to locality, sex and breeding status. Ibis 156, 511–522 (2014).Article 

Google Scholar 
Brasso, R. L. et al. Unique pattern of molt leads to low intraindividual variation in feather mercury concentrations in penguins. Environ. Toxicol. Chem. 32, 2331–2334 (2013).CAS 
PubMed 
Article 

Google Scholar 
Cooper, J. Moult of the black-footed penguin. Int. Zoo Yearb. 18, 22–27 (1978).Article 

Google Scholar 
Cherel, Y., Charrassin, J. & Challet, E. Energy and protein requirements for molt in the king penguin Aptenodytes patagonicus. Am. J. Physiol. 266, R1182–R1188 (1994).CAS 
PubMed 
Article 

Google Scholar 
Brown, C. R. Energetic cost of moult in macaroni penguins (Eudyptes chrysolophus) and rockhopper penguins (E. chrysocome). J. Comp. Physiol. B 155, 515–520 (1985).Article 

Google Scholar 
Dehnhard, N. et al. Survival of rockhopper penguins in times of global climate change. Aquat. Conserv. Mar. Freshw. Ecosyst. 23, 777–789 (2013).
Google Scholar 
Rebstock, G. & Boersma, D. Oceanographic conditions in wintering grounds affect arrival date and body condition in breeding female Magellanic penguins. Mar. Ecol. Prog. Ser. 601, 253–267 (2018).Article 
ADS 

Google Scholar 
Green, J. A., Boyd, I. L., Woakes, A. J., Warren, N. L. & Butler, P. J. Evaluating the prudence of parents: Daily energy expenditure throughout the annual cycle of a free-ranging bird, the macaroni penguin Eudyptes chrysolophus. J. Avian Biol. 40, 529–538 (2009).Article 

Google Scholar 
Crawford, R. J. M., Makhado, A. B., Upfold, L. & Dyer, B. M. Mass on arrival of rockhopper penguins at Marion Island correlated with breeding success. Afr. J. Mar. Sci. 30, 185–188 (2008).Article 

Google Scholar 
Crawford, R. J. M. et al. Food habits of an endangered seabird indicate recent poor forage fish availability off western South Africa. ICES J. Mar. Sci. 76, 1344–1352 (2019).
Google Scholar 
Okes, N. C. et al. Competition for shifting resources in the southern Benguela upwelling: Seabirds versus purse-seine fisheries. Biol. Conserv. 142, 2361–2368 (2009).Article 

Google Scholar 
Campbell, K. J. et al. Local forage fish abundance influences foraging effort and offspring condition in an endangered marine predator. J. Appl. Ecol. 56, 1751–1760 (2019).Article 

Google Scholar 
Grémillet, D. et al. Spatial match-mismatch in the Benguela upwelling zone: Should we expect chlorophyll and sea-surface temperature to predict marine predator distributions?. J. Appl. Ecol. 45, 610–621 (2008).Article 
CAS 

Google Scholar 
Sherley, R. B. et al. Metapopulation tracking juvenile penguins reveals an ecosystem-wide ecological trap. Curr. Biol. 27, 1–6 (2017).Article 
CAS 

Google Scholar 
Sherley, R. B. et al. Influence of local and regional prey availability on breeding performance of African penguins Spheniscus demersus. Mar. Ecol. Prog. Ser. 473, 291–301 (2013).Article 
ADS 

Google Scholar 
Cury, P. M. et al. Global seabird response to forage fish depletion—One-third for the birds. Science 334, 1703–1706 (2011).CAS 
PubMed 
Article 
ADS 

Google Scholar 
Crawford, R. J. M. et al. Collapse of South Africa’s penguins in the early 21st century. Afr. J. Mar. Sci. 33, 139–156 (2011).Article 

Google Scholar 
Sherley, R. B. et al. The conservation status and population decline of the African penguin deconstructed in space and time. Ecol. Evol. 10, 8506–8516 (2020).PubMed 
PubMed Central 
Article 

Google Scholar 
Weller, F. et al. A system dynamics approach to modelling multiple drivers of the African penguin population on Robben Island, South Africa. Ecol. Model. 277, 38–56 (2014).Article 

Google Scholar 
Pichegru, L. Increasing breeding success of an Endangered penguin: Artificial nests or culling predatory gulls?. Bird Conserv. Int. 23, 296–308 (2013).Article 

Google Scholar 
Weller, F. et al. System dynamics modelling of the Endangered African penguin populations on Robben and Dyer islands, South Africa. Ecol. Model. 327, 44–56 (2016).Article 

Google Scholar 
Pichegru, L. et al. Overlap between vulnerable top predators and fisheries in the Benguela upwelling system: Implications for marine protected areas. Mar. Ecol. Prog. Ser. 391, 199–208 (2009).Article 
ADS 

Google Scholar 
Sherley, R. B. et al. Bayesian inference reveals positive but subtle effects of experimental fishery closures on marine predator demographics. Proc. R. Soc. B 285, 20172443 (2018).PubMed 
PubMed Central 
Article 

Google Scholar 
Pichegru, L., Grémillet, D., Crawford, R. J. M. & Ryan, P. G. Marine no-take zone rapidly benefits endangered penguin. Biol. Lett. 6, 498–501 (2010).CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Weller, F. et al. Penguins’ perilous conservation status calls for complementary approach based on sound ecological principles: Reply to Butterworth et al. (2015). Ecol. Model. 337, 1–3 (2016).Article 

Google Scholar 
Butterworth, D. S., Plagányi, E. E., Robinson, W. M. L., Moosa, N. & de Moor, C. L. Penguin modelling approach queried. Ecol. Model. 316, 78–80 (2015).Article 

Google Scholar 
Pichegru, L. et al. Sex-specific foraging behaviour and a field sexing technique for Endangered African penguins. Endanger. Species Res. 19, 255–264 (2013).Article 

Google Scholar 
Roberts, J. African Penguin (Spheniscus demersus) Distribution During the Non-breeding Season: Preparation for, and Recovery from, a Moulting Fast (University of Cape Town, 2016).
Google Scholar 
Dias, M. P. et al. Identification of marine Important Bird and Biodiversity Areas for penguins around the South Shetland Islands and South Orkney Islands. Ecol. Evol. 8, 10520–10529 (2018).PubMed 
PubMed Central 
Article 

Google Scholar 
Lascelles, B. G. et al. Applying global criteria to tracking data to define important areas for marine conservation. Divers. Distrib. 22, 422–431 (2016).Article 

Google Scholar 
Department of Forestry, Fisheries and Environment, T. National data and information report for marine spatial planning: Knowledge baseline for marine spatial planning in South Africa. (2021).Kirkman, S. P. et al. Evaluating the evidence for ecological effectiveness of South Africa’s marine protected areas. Afr. J. Mar. Sci. 43, 389–412 (2021).Article 

Google Scholar 
Harris, L. R. et al. Practical marine spatial management of ecologically or biologically significant marine areas: Emerging lessons from evidence-based planning and implementation in a developing-world context. Front. Mar. Sci. 9, 831678 (2022).Article 

Google Scholar 
Whitehead, T. O., Kato, A., Ropert-Coudert, Y. & Ryan, P. G. Habitat use and diving behaviour of macaroni Eudyptes chrysolophus and eastern rockhopper E. chrysocome filholi penguins during the critical pre-moult period. Mar. Biol. 163, 19 (2016).Article 

Google Scholar 
Warwick-Evans, V., Downie, R., Santos, M. & Trathan, P. N. Habitat preferences of Adélie Pygoscelis adeliae and Chinstrap Penguins Pygoscelis antarctica during pre-moult in the Weddell Sea (Southern Ocean). Polar Biol. 42, 703–714 (2019).Article 

Google Scholar 
Green, C.-P. et al. The role of allochrony in influencing interspecific differences in foraging distribution during the non-breeding season between two congeneric crested penguin species. PLoS ONE 17, e0262901 (2022).CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Pütz, K., Ingham, R. J. & Smith, J. G. Satellite tracking of the winter migration of Magellanic Penguins Spheniscus magellanicus breeding in the Falkland Islands. Ibis 142, 614–622 (2000).Article 

Google Scholar 
Pütz, K. et al. Post-moult movements of sympatrically breeding Humboldt and Magellanic Penguins in south-central Chile. Glob. Ecol. Conserv. 7, 49–58 (2016).Article 

Google Scholar 
Pütz, K., Ingham, R. J., Smith, J. G. & Lüthi, B. H. Winter dispersal of rockhopper penguins Eudyptes chrysocome from the Falkland Islands and its implications for conservation. Mar. Ecol. Prog. Ser. 240, 273–284 (2002).Article 
ADS 

Google Scholar 
Thiebot, J.-B., Cherel, Y., Trathan, P. N. & Bost, C. A. Coexistence of oceanic predators on wintering areas explained by population-scale foraging segregation in space or time. Ecology 93, 122–130 (2012).PubMed 
Article 

Google Scholar 
Thiebot, J.-B., Bost, C.-A., Poupart, T. A., Filippi, D. & Waugh, S. M. Extensive use of the high seas by Vulnerable Fiordland Penguins across non-breeding stages. J. Ornithol. 161, 1033–1043 (2020).Article 

Google Scholar 
Mattern, T. et al. Marathon penguins—Reasons and consequences of long-range dispersal in Fiordland penguins/Tawaki during the pre-moult period. PLoS ONE 13, e0198688 (2018).PubMed 
PubMed Central 
Article 
CAS 

Google Scholar 
Bost, C.-A., Thiebot, J.-B., Pinaud, D., Cherel, Y. & Trathan, P. N. Where do penguins go during the inter-breeding period? Using geolocation to track the winter dispersion of the macaroni penguin. Biol. Lett. 5, 473–476 (2009).CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Baylis, A. M. M., Tierney, M., Orben, R. A., González de la Peña, D. & Brickle, P. Non-breeding movements of gentoo penguins at the Falkland Islands. Ibis 163, 507–518 (2021).Article 

Google Scholar 
Orgeret, F. et al. Exploration during early life: Distribution, habitat and orientation preferences in juvenile king penguins. Mov. Ecol. 7, 29 (2019).CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Thiebot, J. B., Lescroël, A., Barbraud, C. & Bost, C. A. Three-dimensional use of marine habitats by juvenile emperor penguins Aptenodytes forsteri during post-natal dispersal. Antarct. Sci. 25, 536–544 (2013).Article 
ADS 

Google Scholar 
Pütz, K. et al. Post-fledging dispersal of king penguins (Aptenodytes patagonicus) from two breeding sites in the South Atlantic. PLoS ONE 9, e97164 (2014).PubMed 
PubMed Central 
Article 
ADS 
CAS 

Google Scholar 
Birt, V., Birt, T., Goulet, D., Cairns, D. & Montevecchi, W. Ashmole’s halo: Direct evidence for prey depletion by a seabird. Mar. Ecol. Prog. Ser. 40, 205–208 (1987).Article 
ADS 

Google Scholar 
Furness, R. W. & Birkhead, T. R. Seabird colony distributions suggest competition for food supplies during the breeding season. Nature 311, 655–656 (1984).Article 
ADS 

Google Scholar 
Carpenter-Kling, T. et al. Foraging in a dynamic environment: Response of four sympatric sub-Antarctic albatross species to interannual environmental variability. Ecol. Evol. 10, 11277–11295 (2020).PubMed 
PubMed Central 
Article 

Google Scholar 
Kowalczyk, N. D., Reina, R. D., Preston, T. J. & Chiaradia, A. Environmental variability drives shifts in the foraging behaviour and reproductive success of an inshore seabird. Oecologia 178, 967–979 (2015).PubMed 
Article 
ADS 

Google Scholar 
Machovsky-Capuska, G. E. et al. The nutritional nexus: Linking niche, habitat variability and prey composition in a generalist marine predator. J. Anim. Ecol. 87, 1286–1298 (2018).PubMed 
Article 

Google Scholar 
Hays, G. C. et al. Translating marine animal tracking data into conservation policy and management. Trends Ecol. Evol. 34, 459–473 (2019).PubMed 
Article 

Google Scholar 
Kappes, M. A. et al. Hawaiian albatrosses track interannual variability of marine habitats in the North Pacific. Prog. Oceanogr. 86, 246–260 (2010).Article 
ADS 

Google Scholar 
Bost, C. A. et al. Large-scale climatic anomalies affect marine predator foraging behaviour and demography. Nat. Commun. 6, 8220 (2015).CAS 
PubMed 
Article 
ADS 

Google Scholar 
Brown, C. J. et al. Effects of climate-driven primary production change on marine food webs: Implications for fisheries and conservation. Glob. Chang. Biol. 16, 1194–1212 (2010).Article 
ADS 

Google Scholar 
Beever, E. A. et al. Behavioral flexibility as a mechanism for coping with climate change. Front. Ecol. Environ. 15, 299–308 (2017).Article 

Google Scholar 
McInnes, A. M., Ryan, P. G., Lacerda, M. & Pichegru, L. Targeted prey fields determine foraging effort thresholds of a marine diver: Important cues for the sustainable management of fisheries. J. Appl. Ecol. 56, 2206–2215 (2019).Article 

Google Scholar 
van Eeden, R., Reid, T., Ryan, P. G. & Pichegru, L. Fine-scale foraging cues for African penguins in a highly variable marine environment. Mar. Ecol. Prog. Ser. 543, 257–271 (2016).Article 
ADS 

Google Scholar 
Coetzee, J. C., van der Lingen, C. D., Hutchings, L. & Fairweather, T. P. Has the fishery contributed to a major shift in the distribution of South African sardine?. ICES J. Mar. Sci. 65, 1676–1688 (2008).Article 

Google Scholar 
Blamey, L. K. et al. Ecosystem change in the southern Benguela and the underlying processes. J. Mar. Syst. 144, 9–29 (2015).Article 

Google Scholar 
Roy, C., Van Der Lingen, C. D., Coetzee, J. C. & Lutjeharms, J. R. E. Abrupt environmental shift associated with changes in the distribution of Cape anchovy Engraulis encrasicolus spawners in the southern Benguela. Afr. J. Mar. Sci. 29, 309–319 (2007).Article 

Google Scholar 
McInnes, A. M. et al. Small pelagic fish responses to fine-scale oceanographic conditions: Implications for the endangered African penguin. Mar. Ecol. Prog. Ser. 569, 187–203 (2017).CAS 
Article 
ADS 

Google Scholar 
Barange, M., Hampton, I. & Roel, B. A. Trends in the abundance and distribution of anchovy and sardine on the South African continental shelf in the 1990s, deduced from acoustic surveys. S. Afr. J. Mar. Sci. 21, 367–391 (1999).Article 

Google Scholar 
Hutchings, L. et al. Spawning on the edge: Spawning grounds and nursery areas around the southern African coastline. Mar. Freshw. Res. 53, 307–318 (2002).Article 

Google Scholar 
Verheye, H. M., Hutchings, L., Huggett, J. A. & Painting, S. J. Mesozooplankton dynamics in the Benguela ecosystem, with emphasis on the herbivorous copepods. S. Afr. J. Mar. Sci. 12, 561–584 (1992).Article 

Google Scholar 
Hutchings, L., Jarre, A., Lamont, T., van den Berg, M. & Kirkman, S. P. St Helena Bay (southern Benguela) then and now: Muted climate signals, large human impact. Afr. J. Mar. Sci. 34, 559–583 (2012).Article 

Google Scholar 
Goschen, W. S. & Schumann, E. H. Upwelling and the occurrence of cold water around Cape Recife, Algoa Bay, South Africa. S. Afr. J. Mar. Sci. 16, 57–67 (1995).Article 

Google Scholar 
Hutchings, L. et al. The Benguela Current: An ecosystem of four components. Prog. Oceanogr. 83, 15–32 (2009).Article 
ADS 

Google Scholar 
Goschen, W. S., Schumann, E. H., Bernard, K. S., Bailey, S. E. & Deyzel, S. H. P. Upwelling and ocean structures off Algoa Bay and the south-east coast of South Africa. Afr. J. Mar. Sci. 34, 525–536 (2012).Article 

Google Scholar 
van der Lingen, C. D. Diet of sardine Sardinops sagax in the southern Benguela upwelling ecosystem. S. Afr. J. Mar. Sci. 24, 301–316 (2002).Article 

Google Scholar 
van der Lingen, C. D., Hutchings, L. & Field, J. G. Comparative trophodynamics of anchovy Engraulis encrasicolus and sardine Sardinops sagax in the southern Benguela: Are species alternations between small pelagic fish trophodynamically mediated?. Afr. J. Mar. Sci. 28, 465–477 (2006).Article 

Google Scholar 
Wright, K. L. B., Pichegru, L. & Ryan, P. G. Penguins are attracted to dimethyl sulphide at sea. J. Exp. Biol. 214, 2509–2511 (2011).PubMed 
Article 

Google Scholar 
Hagen, C. et al. Evaluating the state of knowledge on fishing exclusions around major African Penguin colonies. (2014).Fort, J. et al. Multicolony tracking reveals potential threats to little auks wintering in the North Atlantic from marine pollution and shrinking sea ice cover. Divers. Distrib. 19, 1322–1332 (2013).Article 

Google Scholar 
Reiertsen, T. K. et al. Prey density in non-breeding areas affects adult survival of black-legged kittiwakes Rissa tridactyla. Mar. Ecol. Prog. Ser. 509, 289–302 (2014).Article 
ADS 

Google Scholar 
Fayet, A. L. et al. Ocean-wide drivers of migration strategies and their influence on population breeding performance in a declining seabird. Curr. Biol. 27, 3871–3878 (2017).CAS 
PubMed 
Article 

Google Scholar 
Desprez, M., Jenouvrier, S., Barbraud, C., Delord, K. & Weimerskirch, H. Linking oceanographic conditions, migratory schedules and foraging behaviour during the non-breeding season to reproductive performance in a long-lived seabird. Funct. Ecol. 32, 2040–2053 (2018).Article 

Google Scholar 
Randall, R. M. & Randall, B. The annual cycle of the Jackass Penguin Spheniscus demersus at St Croix Island, South Africa. In Proc. Symp. Birds Sea Shore 427–450 (1981).Wolfaardt, A. C., Underhill, L. G. & Visagie, J. Breeding and moult phenology of African penguins Spheniscus demersus at Dassen Island. Afr. J. Mar. Sci. 31, 119–132 (2009).Article 

Google Scholar 
Crawford, R. J. M. et al. Molt of the African penguin, Spheniscus demersus, in relation to its breeding season and food availability. Acta Zool. Sin. 52, 444–447 (2006).
Google Scholar 
Randall, R. M. Biology of the Jackass Penguin Spheniscus demersus (L.) at St Croix, South Africa (Univeristy of Port Elizabeth, 1983).
Google Scholar 
Harding, C. T. Tracking African Penguins (Spheniscus demersus) Outside of the Breeding Season: Regional Effects and Fishing Pressure During the Pre-moult Period (University of Cape Town, 2013).
Google Scholar 
Wilson, R. P. The Jackass Penguin (Spheniscus demersus) as a pelagic predator. Mar. Ecol. Prog. Ser. 25, 219–227 (1985).Article 
ADS 

Google Scholar 
Freitas, C. argosfilter: Argos locations filter. (2012).Worton, B. J. Kernel methods for estimating the utilization distribution in home-range studies. Ecology 70, 164–168 (1989).Article 

Google Scholar 
Calenge, C. The package ‘adehabitat’ for the R software: A tool for the analysis of space and habitat use by animals. Ecol. Modell. 197, 516–519 (2006).Article 

Google Scholar 
Vander Wal, E. & Rodgers, A. R. An individual-based quantitative approach for delineating core areas of animal space use. Ecol. Model. 224, 48–53 (2012).Article 

Google Scholar 
Dinno, A. dunn.test: Dunn’s Test of Multiple Comparisons Using Rank Sums (2017).Bhattacharyya, A. On a measure of divergence between two multinomial populations. Indian J. Stat. 7, 401–406 (1946).MathSciNet 
MATH 

Google Scholar 
Beal, M. et al. track2KBA: An R package for identifying important sites for biodiversity from tracking data. Methods Ecol. https://doi.org/10.1111/2041-210X.13713 (2021).Article 

Google Scholar 
Donald, P. F. et al. Important Bird and Biodiversity Areas (IBAs): The development and characteristics of a global inventory of key sites for biodiversity. Bird Conserv. 29, 177–198 (2019).Article 

Google Scholar 
Handley, J. M. et al. Evaluating the effectiveness of a large multi-use MPA in protecting Key Biodiversity Areas for marine predators. Divers. Distrib. 26, 715–729 (2020).Article 

Google Scholar 
Strimas-Mackey, M. smoothr: Smooth and tidy spatial features. R package version 0.2.2. https://CRAN.R-project.org/package=smoothr (2018).Department of Forestry Fisheries and the Environment, T. South Africa Marine Protected Area Zonations (SAMPAZ_OR_2021_Q3). https://egis.environment.gov.za/data_egis/data_dow (2021).R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing https://www.r-project.org/ (2021). More

Trait modelsOur analysis included 491,001 unique trait measurements across 18 traits, encompassing 13,189 tree species from 2313 genera, reflecting ~21% of all known tree species33 (Fig. 1). Traits were measured at 8683 locations across the globe and 373 distinct eco-regions (Supplementary Tables 1, 2), with georeferenced measurements capturing 15% of known tree species in Eurasia, 13% in South America, 9% in Oceania, and 6% in North America and Africa33. The raw data covered 22% of all trait-by-species combinations (Fig. 1b, Supplementary Fig. 2), nearly identical to other large-scale trait analyses across the entire plant kingdom5,17,30. Yet there was considerable variation in coverage across traits, with traits such as specific leaf area and leaf nitrogen measured on more than 60% of all species, versus traits such as crown diameter and conduit diameter, which captured fewer than 5% of species (Fig. 1b, Supplementary Fig. 2). Across all species, 423 had more than 10 unique traits measured, and two species (Picea abies and Pinus sylvestris) had measurements for all 18 traits. In general, there was highly consistent coverage across taxonomic orders and traits (Supplementary Fig. 1), with gymnosperms being slightly overrepresented (comprising 3.1 ± 6.8% of measurements in the database versus ~1% of all known tree species34,35, Fig. 1a), in part reflecting the wider geographic range of many gymnosperms relative to angiosperms36.To explore relationships in functional traits at the individual level, we used random-forest machine-learning models to estimate missing trait values for each individual tree as a function of its environment and phylogenetic history. We also conducted a second set of analyses where trait expression was estimated using phylogenetic information only, which allowed us to include additional non-georeferenced data (Fig. 1), while also quantifying the relative contribution of environmental information on trait expression (Supplementary Fig. 6). Following standard approaches5,15,29,30, all traits were log-transformed and standardized to allow for statistically robust comparisons. Environmental predictors included ten variables encompassing climate37,38,39,40, soil41, topographic42, and geological43 features. Phylogenetic history was incorporated via the first ten phylogenetic eigenvectors44,45 (see Methods). By including environmental information alongside phylogenetic information, this approach not only allowed us to impute species-level traits which have strong phylogenetic signals and weak environmental signals, as is traditionally done17,30 but also to robustly estimate traits which have a weak phylogenetic signal and are instead strongly sensitive to environmental conditions. Moreover, being a non-parametric approach, the random forest makes no a priori assumptions about how trait expression varies across phylogenetic groups or environments.Across all 18 traits, the best-fitting models explained 54 ± 14% of out-of-fit trait variation (VEcv, see Methods), ranging from 26% for stem diameter to 76% of the variation in leaf area (Supplementary Figs. 6, 7). This accuracy was quantified using buffered leave-one-out cross-validation to account for spatial and phylogenetic autocorrelation46, and thus serves as a conservative lower bound for species which are phylogenetically and environmentally distinct from the observations47. There was no significant relationship between out-of-fit cross-validation accuracy and sample size (R2 = 0.06, p = 0.33), highlighting the relatively broad taxonomic coverage for each trait (Fig. 1, Supplementary Fig. 1).Environmental variables and phylogenetic information had approximately equal explanatory power (relative importance of 0.51 vs 0.49 for environment vs. phylogeny), albeit with substantial variation across traits (Supplementary Fig. 9). The inclusion of environmental variables increased the explanatory power of the models by 35%, on average (Supplementary Fig. 6), with crown diameter, crown height, leaf density, and stem diameter exhibiting the largest relative increases (54%, 45%, 73%, and 26%, respectively), mirroring the fact that these traits have comparatively low phylogenetic signal relative to other traits (assessed via Pagel’s λ on the raw data, Fig. 4c). Seed dry mass was the only trait with a substantial increase in accuracy using the phylogeny-only model (25% improvement; Supplementary Fig. 6), reflecting the fact that seed dry mass had the strongest phylogenetic signal of all traits (Fig. 4c), and also because this trait has a substantial amount of additional non-georeferenced data that was included in the phylogeny-only models (Fig. 1b). Wood density was the only trait with nearly identical predictive power whether or not environmental information was included, whereas all other traits exhibited significantly reduced accuracy when environmental information was excluded (Supplementary Fig. 6).Relationships in tree trait expressionUsing the resulting trait models, we imputed missing trait values for every tree with at least one georeferenced trait measurement. For all traits except seed dry mass, we used the random-forest models accounting for environmental and phylogenetic information; for seed dry mass, we used the phylogeny-only model to estimate expression due to its substantially higher data availability and out-of-fit accuracy. For tree height, stem diameter, crown height, crown width, and root depth, we used quantile random forest48 to estimate the upper 90th percentile value for each species in its given location, thereby minimizing ontogenetic variation across a tree’s lifetime (see Methods). We used the resulting trait data to explore the dominant drivers of trait variation using species-weighted principal component analysis, accounting for an unequal number of observations across species.When considering all traits simultaneously, the first two axes of the resulting principal components (PC) capture 41% of the variation in overall trait expression (Fig. 2a; Supplementary Fig. 10; Supplementary Table 5). The first trait axis correlates most strongly with leaf thickness, specific leaf area, and leaf nitrogen (PC loadings of L = 0.77, 0.74, and 0.73, respectively). By capturing key aspects of the leaf-economic spectrum14, these traits reflect various physiological controls on leaf-level resource processing, tissue turnover and photosynthetic rates49. Thick leaves with low specific leaf area (SLA) can help minimize desiccation, frost damage, and nutrient limitation, but at the cost of reduced photosynthetic potential due to primary investment in structural resistance50. Accordingly, leaf nitrogen—a crucial component of Rubisco for photosynthesis51—trades off strongly with leaf thickness. This first axis thus captures the core distinction between “acquisitive” (fast) and “conservative” (slow) life-history strategies across the plant kingdom7,52, reflecting an organismal-level trade-off between the high photosynthetic potential in optimal conditions versus abiotic tolerance in suboptimal conditions. Nevertheless, leaf density—which is related to SLA and is a key feature of the leaf-economic spectrum—loads relatively weakly on this first trait axis compared to other leaf traits (L = −0.28 for axis 1, vs 0.20 for axis 2; Supplementary Table 5), highlighting important aspects of leaf structure that are not captured by this dominant trait axis53.Fig. 2: The dominant trait axes and relationships.Shown are the first two principal component axes capturing trait relationships across the 18 functional traits. a All tree species (n = 30,146 observations), b angiosperms only (n = 24,658), and c gymnosperms only (n = 5498). In a the three variables that load most strongly on each axis are shown in dark black lines, with the remaining variables shown in light grey. These same six variables are highlighted in b and c illustrating how the same relationships extend to angiosperms and gymnosperms (see Supplementary Figs. 10–12 for the full PCAs with all traits visible, and Supplementary Table 5 for the PC loadings).Full size imageThe second trait axis correlates most strongly with maximum tree height (PC loading of L = 0.77), crown height, (L = 0.75), and crown diameter (L = 0.88), highlighting the overarching importance of competition for light and canopy position in forests7 (Fig. 2a; Supplementary Fig. 10; Supplementary Table 5). Large trees and large crowns are critical for light access and for maximizing light interception down through the canopy54. Nevertheless, tall trees with deep crowns also experience greater susceptibility to disturbance and mechanical damage, primarily due to wind and weight25. Because of the massive carbon and nutrient costs required to create large woody structures55,56, larger trees are less viable in nutrient-limited or colder climates57, and in exposed areas with high winds or extreme weather events58. This second axis thus reflects a fundamental biotic/abiotic trade-off related to overall tree size, which is largely orthogonal to leaf-level nutrient-use and photosynthetic capacity.Despite substantial differences in wood and leaf structures between angiosperms and gymnosperms (e.g. vessels vs. tracheids), the two main relationships hold within, as well as across, angiosperms and gymnosperms (Fig. 2b, c; Supplementary Figs. 11, 12). Indeed, angiosperms and gymnosperms are subject to the same physical, mechanical, and chemical processes that determine the ability to withstand various biotic and abiotic pressures59.Collectively, these two primary trait axes capture two dominant ecological trade-offs that underpin tree survival in any given environment: (1) the ability to maximize leaf photosynthetic activity, at the cost of increased risk of leaf desiccation, and (2) the ability to compete for space and maximize light interception, at the cost of increased susceptibility to mechanical damage. By capturing two aspects of conservative-acquisitive life-history strategies, these two relationships closely mirror those seen when considering herbaceous species alongside woody species5,17. However, in line with our expectations, these two axes capture only ~40% of the variation in trait space, versus nearly ~75% of variation when considering only six traits across the entire plant kingdom5. Here, the first seven PC axes are needed to account for 75% of the variation across all 18 traits (Supplementary Table 5). Thus, while this analysis supports the universality of these two primary PC axes, it also demonstrates that the majority of trait variation in trees is unexplained by these two dimensions. As such, quantifying the full dimensionality of trait space by exploring multidimensional trait clusters is needed to better capture the wide breadth of tree form and function.Environmental predictors of trait relationshipsTo examine how environmental variation shapes trait expression across the globe, we next quantified the relationships between environmental conditions and the dominant trait axes. Using Shapley values60, we partitioned the relative influence of each environmental variable on the PC trait axes, controlling for all other variables in the model (see Methods).In line with previous analysis across the plant kingdom61, temperature variables were the strongest drivers of trait relationships (Fig. 3, Supplementary Figs. 17, 18), with annual temperature having the strongest influence both on leaf-economic traits (PC axis 1, Fig. 3c) and on tree-size traits (PC axis 2, Fig. 3d). Leaves face increased frost risk and reduced photosynthetic potential in colder conditions, such that ecological selection should favour thick leaves with low SLA over thin leaves with high SLA and high nutrient-use49. Trees in warm environments are more likely to experience strong biotic interactions, which should increase evolutionary and ecological selection pressures over time62,63, favouring tall species with large crowns that have high competitive ability and efficient light acquisition strategies. Annual temperature thus predominantly reflects the transition from gymnosperm- to angiosperm-dominated ecosystems, with this inflection point occurring at ~15 °C for both axes, demonstrating strong environmental convergence between the dominant axes of trait variation.Fig. 3: The relationship between environmental variables and trait axes.a, b The relative influence of the environmental variables on the two dominant PC axes. The ten variables are sorted by overall variable importance in the models (see Methods). Yellow points are observations which have high values of that environmental variable; blue values are the lowest. Points to the right of zero indicate a positive influence on the PC axis; points to the left indicate a negative influence (see also Supplementary Figs. 17, 18). c–h The relationships between environmental variables and PC axis values for the three variables in a with the strongest influence. Values above zero show a positive influence on PC axis values; values less than zero indicate a negative influence.Full size imageBeyond annual temperature, each trait axis demonstrated different relationships with climate, soil, and topographic variables (Fig. 3a, b, Supplementary Figs. 17, 18). Percent sand content had the second-highest influence on the first trait axis (Fig. 3e), supporting patterns seen across the entire plant kingdom17. Sand content is a strong proxy for soil moisture and soil-available nutrients such as phosphorous, and is therefore closely tied to leaf photosynthetic rates64. In contrast to previous work, however, we find that soil characteristics have correspondingly little effect on the second axis of trait variation (Fig. 3b; Supplementary Fig. 18). Instead, precipitation was the second strongest driver of tree height and crown size (Fig. 3f), with large trees with large crowns becoming consistently more frequent with increasing precipitation. These results highlight that, despite the primary importance of temperature, the main climate stressors to trees (e.g. xylem cavitation and embolism, fire regimes, and leaf desiccation) typically arise via interactions between temperature, soil nutrients, and water availability.For both axes, elevation was the third strongest driver of trait values (Fig. 3g, h), highlighting a critical component of tree functional biogeography that extends beyond climate and soil. Yet the effects of elevation on trait expression differed somewhat across the two axes. For the first axis related to leaf-economic traits, there is little influence at low elevations, followed by a sharp transition at ~2000 m towards gymnosperm-dominated species with thick leaves, low SLA, and low leaf N. For the second trait axis related to tree size, elevation instead has a strong positive influence on tree height and crown size at low elevations, which becomes increasingly less influential past ~500 m. Such results partly reflect the transition from angiosperm to gymnosperm-dominated stands at higher elevations (blue vs. red points, Fig. 3g, h), and potentially the role of environmentally mediated intraspecific variation in traits such as tree height65,66.These results demonstrate close alignment of the dominant trait PC axes across biogeographic regions. Despite the orthogonality of these axes in trait species, environmental conditions place similar constraints on both trait axes, particularly at the environmental extremes (e.g. warm, moist, low elevation vs. cold, dry, high elevation), leading to convergence of the dominant trait axes across environmental gradients.Trait clusters at the global scaleTo better explore the multidimensional nature of trait relationships that are not fully covered by the dominant two axes, we subsequently identified groups of traits that form tightly coupled clusters and which reflect distinct aspects of tree form and function.Our results show that these 18 traits can be grouped into eight trait clusters, each of which reflects a unique aspect of morphology, physiology, or ecology (Fig. 4a, Supplementary Fig. 23). The largest trait cluster (Fig. 4a, pink cluster) demonstrates wood/leaf integration of moisture regulation and photosynthetic activity via the inclusion of leaf area, stem conduit diameter, stomatal conductance, and leaf Vcmax (the maximum rate of carboxylation). Distinct from this cluster are the three traits loading most strongly on PC axis 1 (SLA, leaf thickness and leaf N; Fig. 4a, yellow), highlighting complementary aspects of the leaf-economic spectrum indicative of acquisitive vs. conservative resource use15. The role of leaf K and P in leaf nutrient economies are well established7,67, and yet these traits form a distinct cluster from the other leaf-economic traits (Fig. 4a, light blue) due to their relatively high correlation with tree height and crown size, particularly for leaf K, which loads almost equally on both trait axes (Fig. 4b, Supplementary Table 5).Fig. 4: Trait correlations and functional clusters.a Trait clusters with high average intra-group correlation. The upper triangle gives the species-weighted correlations incorporating intraspecific variation. The lower triangle gives the corresponding correlations among phylogenetic independent contrasts, which adjusts for pseudo-replication due to the non-independence of closely related species. The size of the circle denotes the relative strength of the correlation, with solid circles denoting positive correlations and open circles denoting negative correlations (see Supplementary Fig. 19 for the numeric values). b PC loadings for each trait and each of the first two principal component axes, illustrating which functional trait clusters align most strongly with the dominant axes of trait variation (see Supplementary Table 5 for the full set of PC loadings). c The species-level phylogenetic signal of each trait (Pagel’s λ), calculated using only the raw trait values.Full size imageTree height and crown size form their own distinct cluster (Fig. 4a, dark green), further supporting the inference that these traits reflect key aspects of tree form and function independent of the leaf-economic spectrum. Yet leaf area, despite being part of the cluster reflecting moisture regulation and photosynthetic activity, loads almost equally on PC axes 1 and 2 (Fig. 4b, Supplementary Table 5), highlighting that it serves as an intermediary between the two key aspects of tree size and leaf economics. It is a critical driver of moisture regulation and photosynthetic capacity, while also playing an important role in the light acquisition, leaf-turnover time, and competitive ability54,68.There are two additional two-trait clusters, both of which load relatively poorly on the two primary PC axes: (1) stem diameter and bark thickness (Fig. 4, dark blue), and (2) wood and leaf density (Fig. 4, light green). Bark thickness increases with tree size not only as a result of bark accumulation as trees age, but also due to the functional/metabolic needs of the plant69,70. From an ecological perspective, thick bark can be critical for defense against fire and pest damage (mainly a thick outer bark region), for storage and photosynthate transportation needs (mainly a thick inner bark region)71,72. Yet such relationships are strongly ecosystem-dependent, with tree size emerging as the dominant driver at the global scale70. In contrast, wood density and leaf density are strongly linked to slow/fast life-history strategies, where denser plant parts reduce growth rate and water transport6,15 but protect against pest damage, desiccation, and mechanical breakage6,50,56. As such, leaf density captures fundamentally unique aspects of leaf form and function relative to other leaf traits such as SLA53 (Fig. 4b, Supplementary Table 5), and our results support the inference that these translate into fundamentally different ecological strategies73. Collectively, these two-trait clusters each demonstrate unique and complementary mechanisms that insulate trees against various disturbances and extreme weather events, but at the cost of reduced growth, competitive ability, and productivity under optimal conditions (see Supplementary Notes).Lastly, two traits each comprise their own unique cluster: root depth and seed dry mass (Fig. 4a, purple and orange, respectively). Root growth is subject to a range of belowground processes (e.g. root herbivory, depth to bedrock), and our results confirm previous work demonstrating a clear disconnect between aboveground and belowground traits23,74,75. Root depth accordingly has a relatively weak phylogenetic signal (λ = 0.44, Fig. 4c) but a strong environmental signal (Supplementary Figs. 6, 9), reflecting distinct belowground constraints on trait expression23. In contrast, seed dry mass exhibits the strongest phylogenetic signal (λ = 0.98, Fig. 4c) and weakest environmental signal of any trait (Supplementary Figs. 6, 9), and it accordingly was the only trait where the phylogeny-only model performed substantially better (Supplementary Fig. 6). In line with previous work, seed dry mass has moderate correlations with various other traits underpinning leaf economics and tree size5,28 (e.g. ρ = 0.28, −0.22, and 0.22 for tree height, leaf K, and leaf density, using the raw data), yet it exhibits relatively weak correlation with most other traits, placing it in a distinct functional cluster. Reproductive traits are subject to unique evolutionary pressures26, indicative of different seed dispersal vectors (wind, water, animals) and various ecological stressors that uniquely affect seed viability and germination26. The emergence of root depth and seed dry mass as solo functional clusters thus supports the previous inference that belowground traits74 and reproductive traits26 reflect distinct aspects of tree form and function not fully captured by leaf or wood trait spectrums. More