Ecology

1.Mayaux, P. et al. Tropical forest cover change in the 1990s and options for future monitoring. Philos. Trans. R. Soc. Lond. B Biol. Sci. 60(1454), 373–384. https://doi.org/10.1098/rstb.2004.1590 (2005).Article 

Google Scholar 
2.Lovejoy, T. E. Biodiversity: What is it? In Biodiversity II: Understanding and Protecting Our Biological Resources (eds Reaka-Kudla, M. L. et al.) 7–14 (Joseph Henry Press, 1997).
Google Scholar 
3.Harris, L. D. The Fragmented Forest: Island Biogeographic Theory and the Preservation of Biological Diversity (The University of Chicago Press, 1984).Book 

Google Scholar 
4.Achard, F. et al. Determination of deforestation rates of the world’s humid tropical forests. Science 297(5583), 999–1002. https://doi.org/10.1126/science.1070656 (2002).ADS 
CAS 
Article 
PubMed 

Google Scholar 
5.NRSA – 1983. Mapping of forest cover in India from satellite imagery (1972–75 and 1980–82). Summary Report, National Remote Sensing Agency, Hyderabad, India, pp. 5–6.6.FAO – 2000. Global forest resources assessment. Chapter 23. South Asia. Food and Agriculture Organization, Rome, Italy. http://www.fao.org/3/Y1997E/y1997e0s.htm#bm28. (Accessed on 8 August 2020).7.Datta, D. & Deb, S. Analysis of coastal land use/land cover changes in the Indian Sunderbans using remotely sensed data. Geo-sp. Inf. Sci. 15, 241–250. https://doi.org/10.1080/10095020.2012.714104 (2012).Article 

Google Scholar 
8.Reddy, C. S., Jha, C. S. & Dadhwal, V. K. Assessment and monitoring of long-term forest cover changes in Odisha, India using remote sensing and GIS. Environ. Monit. Assess. 185, 4399–4415. https://doi.org/10.1007/s10661-012-2877-5 (2013).Article 
PubMed 

Google Scholar 
9.Champion, H. G. & Seth, S. K. A revised forest types of India (Manager of Publications, 1968).
Google Scholar 
10.FSI – 2019. State of forest report, Assam. Forest Survey of India, Ministry of Environment and Forests, Dehradun, pp. 23–33.11.Assam Times – 2019. Encroachment killing forest in the state. https://www.assamtimes.org/node/22026. (Accessed on 25 August 2020).12.Saikia, A., Hazarika, R. & Sahariah, D. Land-use/land-cover change and fragmentation in the Nameri Tiger Reserve India. Danish J. Geogr. 113(1), 1–10. https://doi.org/10.1080/00167223.2013.782991 (2013).Article 

Google Scholar 
13.Assam Human Development Report – 2014. Managing diversities, achieving human development. Omeo Kumar Das Institute of Social Change and Development and Institute for Human Development, Planning and Development Department, Government of Assam. https://niti.gov.in/writereaddata/files/human-development/Assam_HDR_30Sep2016.pdf. (Accessed on 26 August 2020).14.Woods, C. H. & Skole, D. Linking satellite, census, and survey data to study deforestation in the Brazilian Amazon. In People and Pixels: Linking Remote Sensing and Social Science (eds Liverman, D. et al.) 70–90 (National Academy Press, 1998). https://doi.org/10.17226/5963.
Google Scholar 
15.Srivastava, S., Singh, T. P., Singh, H., Kushwaha, S. P. S. & Roy, P. S. Assessment of large scale deforestation in Sonitpur district of Assam. Curr. Sci. 82, 1480–1484 (2002).
Google Scholar 
16.Harper, G. J., Steininger, M. K., Tucker, C. J., Juhn, D. & Hawkins, F. Fifty years of deforestation and forest fragmentation in Madagascar. Environ. Conserv. 34(4), 1–9. https://doi.org/10.1017/S0376892907004262 (2007).Article 

Google Scholar 
17.Manjula, K. R., Jyothi, S., Varma, A. K. & Kumar, S. V. Construction of spatial dataset from remote sensing using GIS for deforestation study. Int. J. Comput. Appl. 31(10), 26–32 (2011).
Google Scholar 
18.Phukan, P., Thakuriah, G. & Saikia, R. Land use land cover change detection using remote sensing and GIS techniques: A case study of Golaghat district of Assam, India. Int. Res. J. Earth Sci. 1(1), 11–15 (2013).
Google Scholar 
19.Armenta, S. A. M. et al. Determination and analysis of hot spot areas of deforestation using Remote Sensing and Geographic Information System techniques. Case study: State Sinaloa, Mexico. Open J. For. 6, 295–304. https://doi.org/10.4236/ojf.2016.64024 (2016).Article 

Google Scholar 
20.Sarma, P. K. et al. Land-use and land-cover change and future implication analysis in Manas National Park, India using multi-temporal satellite data. Curr. Sci. 95(2), 223–227 (2008).
Google Scholar 
21.Valožić, L. & Cvitanović, M. Mapping the forest change: using Landsat imagery in forest transition analysis within the Medvednica protected area. Hrvat. Geo. Glas. 73(1), 245–255. https://doi.org/10.21861/hgg.2011.73.01.16 (2011).Article 

Google Scholar 
22.Gambo, J., Mohd Shafri, H. Z., Shaharum, N. S., Abidin, F. A. & Rahman, M. T. Monitoring and predicting land use-land cover (LULC) changes within and around Krau wildlife reserve (KWR) protected area in Malaysia using multi-temporal Landsat data. Geoplanning: J. Geomatics Plan. 5(1), 17–34. https://doi.org/10.14710/geoplanning.5.1.17-34 (2018).‬23.Bapu, T. D. & Nimasow, G. Land cover change assessment of Pakke Tiger Reserve (PTR), East Kameng district of Arunachal Pradesh. J. Remote Sens. & GIS, 9(1), 26–33. http://doi.org/https://doi.org/10.37591/.v9i1.93 (2018).24.Kushwaha, S. P. S. & Hazarika, R. Assessment of habitat loss in Kameng and Sonitpur Elephant reserves. Curr. Sci. 87(10), 1447–1453 (2004).
Google Scholar 
25.Census of India – 2011. Primary Census Abstracts. Registrar General of India, Ministry of Home Affairs, Government of India, Retrieved from https://www.censusindia.gov.in/2011census/PCA/pca_highlights/pe_data.html26.Bose, A. U. Tracking the forest rights act in Nameri National Park & Sonai Rupai Wildlife Sanctuary. A report of Kalpavriksh Environmental Action Group, Pune, Maharashtra. https://kalpavriksh.org/wp-content/uploads/2020/07/Assam-Poster_August14_FINAL1.pdf. (2009). (Accessed on 25 August 2020).27.Das, N. Assessment of ecotourism resources: An applied methodology to Nameri National Park of Assam-India. J. Geogr. Reg. Plan. 6(6), 218–228. https://doi.org/10.5897/JGRP12.057 (2013).Article 

Google Scholar 
28.Dong, J. et al. Mapping deciduous rubber plantations through integration of PALSAR and multitemporal landsat imagery. Remote Sens. Environ. 134, 392–402. https://doi.org/10.3390/rs70101048 (2013).ADS 
Article 

Google Scholar 
29.USGS – 2003. Preliminary Assessment of the Value of Landsat 7 ETM+ Data Following Scan Line Corrector Malfunction. USA: EROS Data Center. United States Geological Survey. https://landsat.usgs.gov/sites/default/files/documents/SLC_off_Scientific_Usability.pdf. (Accessed on 8 August 2020).30.Settle, J. J. & Briggs, S. S. Fast maximum likelihood classification of remotely sensed imagery. Int. J. Remote Sens. 8, 723–734. https://doi.org/10.1080/01431168708948683 (1987).ADS 
Article 

Google Scholar 
31.Richards, J. A. Remote Sensing Digital Image Analysis: An introduction. https://doi.org/10.1007/978-3-642-30062-2_8 (Springer, 2013).Book 

Google Scholar 
32.Stehman, S. V. & Czaplewski, R. L. Design and analysis for thematic map accuracy assessment: Fundamental principles. Remote Sens. Environ. 64, 331–334. https://doi.org/10.1016/S0034-4257(98)00010-8 (1998).ADS 
Article 

Google Scholar 
33.Story, M. & Congalton, R. G. Accuracy assessment: A user’s perspective. Photogram. Eng. Rem. S. 52(3), 397–399 (1986).
Google Scholar 
34.Munoz, S. R. & Bangdiwala, S. I. Interpretation of kappa and B statistics measures of agreement. J. Appl. Stat. 24(1), 105–111. https://doi.org/10.1080/02664769723918 (1997).Article 

Google Scholar 
35.Sim, J. & Wright, C. C. The Kappa statistic in reliability studies: Use, interpretation, and sample size requirements. Phys. Ther. 85(3), 257–268 (2005).Article 

Google Scholar 
36.Landis, J. R. & Koch, G. G. The measurement of observer agreement for categorical data. Biometrics 33, 159–174. https://doi.org/10.2307/2529310 (1977).CAS 
Article 
PubMed 
MATH 

Google Scholar 
37.Balasubramanian, D., Arunachalam, K. & Arunachalam, A. Human-induced land use/land-cover change and bioresource management in Bura Chapori Wildlife Sanctuary in North-East India. Clim. Change Environ. Sustain. 4(1), 28–37. https://doi.org/10.5958/2320-642X.2016.00005.3 (2016).Article 

Google Scholar 
38.Sugden, A. M. Mapping global deforestation patterns. Science 361(6407), 1083. https://doi.org/10.1126/science.361.6407.1083-e (2018).Article 

Google Scholar 
39.Curtis, P. G., Slay, C. M., Harris, N. L., Tyukavina, A. & Hansen, M. C. Classifying drivers of global forest loss. Science 361(6407), 1108–1111. https://doi.org/10.1126/science.aau3445 (2018).ADS 
CAS 
Article 
PubMed 

Google Scholar 
40.Myers, N. Tropical deforestation: rates and patterns. In K. Brown & D. Pearce (Eds.), The causes of tropical deforestation. The economic and statistical analysis of factors giving rise to the loss of the tropical forest (pp. 27–40). London: UCL Press (1994).41.Barraclough, S. & Ghimire, K. B. Agricultural Expansion and Tropical Deforestation (Virginia, 2000).
Google Scholar 
42.Hansen, M. C. et al. High-resolution global maps of the 21st-century forest cover change. Science 342, 850–853. https://doi.org/10.1126/science.1244693 (2013).ADS 
CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
43.Reddy, C. S., Rao, P. R. M., Pattanaik, C. & Joshi, P. K. Assessment of large-scale deforestation in Nawarangpur district, Orissa, India: A remote sensing based study. Environ. Monit. Assess. 154, 325–335. https://doi.org/10.1007/s10661-008-0400-9 (2009).Article 

Google Scholar 
44.Saikia, A. Drivers of forest loss. In A. Saikia (Eds.), Over-exploitation of forests. Springer Briefs in Geography. Springer, Cham, Switzerland. https://doi.org/10.1007/978-3-319-01408-1_7 (2014).45.Lele, N. & Joshi, P. K. Analyzing deforestation rates, spatial forest cover changes and identifying critical areas of forest cover changes in North-East India during 1972–1999. Environ. Monit. Assess. 156, 159–170. https://doi.org/10.1007/s10661-008-0472-6 (2009).Article 
PubMed 

Google Scholar 
46.Joppa, L. N., Loarie, S. R. & Pimm, S. L. On the protection of ‘“protected areas”’. Proc. Natl. Acad. Sci. U.S.A. 105(18), 6673–6678 (2008).ADS 
CAS 
Article 

Google Scholar 
47.Bharucha, E. Textbook of Environmental Studies for Undergraduate Courses (Universities Press, 2005).
Google Scholar 
48.Talukdar, N. R. & Choudhury, P. Conserving wildlife wealth of Patharia Hills Reserve Forest, Assam, India: A critical analysis. Glob. Ecol. Conserv. 10, 126–138. https://doi.org/10.1016/j.gecco.2017.02.002 (2017).Article 

Google Scholar 
49.Sonitpur District Judiciary – 2016. In the Court of Additional Sessions Judge, Sonitpur, Tezpur. Sessions Case No. 224 of 2016, U/s. 51 of Wildlife (Protection) Act, 1972. http://sonitpurjudiciary.gov.in/Judgement/09_Sessions%20Case%20No.224%20of%202016.pdf. (Accessed on 2 August 2020.50.Assam Forest Department – 2014. A draft proposal for declaring Eco-Sensitive Zone around Sonai-Rupai Wildlife Sanctuary. Prepared by Divisional Forest Officer Western Assam Wildlife Division, Tezpur, Assam Forest Department, Government of Assam. http://103.8.249.31/assamforest/notificationsOrders/ESZ%20Sonai-Rupai_final.pdf. (Accessed on 25 August 2020.51.ESZ Expert Committee Meeting – 2020. Minutes of 41st ESZ expert committee meeting for the declaration of Eco-Sensitive Zone (ESZ) around protected areas & Zonal Master Plan through video conferencing held on 23rd to 24th June 2020. Ministry of Environment, Forests and Climate Change, Government of India. http://moef.gov.in/wp-content/uploads/2019/10/41-st-ECM_Approved-minutes_.pdf. (Accessed on 25 August 2020). More

1.Møller, A., Fiedler, W. & Berthold, P. Effects of climate change on birds (Oxford University Press, 2010).
Google Scholar 
2.IPCC. Global Warming of 1.5°C. An IPCC Special Report on the impacts of global warming of 1.5°C above pre-industrial levels and related global greenhouse gas emission pathways, in the context of strengthening the global response to the threat of climate change, sustainable development, and efforts to eradicate poverty. (The Intergovernmental Panel on Climate Change, 2018).3.Lovejoy, T. E., Hannah, L. & Wilson, E. O. Biodiversity and climate change (Yale University Press, 2019).Book 

Google Scholar 
4.Hughes, L. Biological consequences of global warming: is the signal already apparent?. Trends Ecol. Evol. 15, 56–61 (2000).CAS 
PubMed 
Article 

Google Scholar 
5.Moore, N. Precipitation regimes and climate change. In Global Environmental Change (ed. Freedman, B.) 191–197 (Springer, Dordrecht, 2014).
Google Scholar 
6.Knapp, A. K. et al. Characterizing differences in precipitation regimes of extreme wet and dry years: Implications for climate change experiments. Glob. Change Biol. 21, 2624–2633 (2015).ADS 
Article 

Google Scholar 
7.Tobias, A. & Díaz, J. Heat waves, human health, and climate change. In Global Environmental Change (ed. Freedman, B.) 447–453 (Springer, Dordrecht, 2014).
Google Scholar 
8.Freedman, B. Global Environmental Change (Springer, 2014).Book 

Google Scholar 
9.Hannah, L. Climate change biology 2nd edn. (Elsevier, 2014).
Google Scholar 
10.Gibbons, J. W. et al. The global decline of reptiles, Déjà Vu Amphibians: Reptile species are declining on a global scale. Six significant threats to reptile populations are habitat loss and degradation, introduced invasive species, environmental pollution, disease, unsustainable use, and global climate change. BioScience 50, 653–666 (2000).11.Williams, S. E., Shoo, L. P., Isaac, J. L., Hoffmann, A. A. & Langham, G. Towards an integrated framework for assessing the vulnerability of species to climate change. PLoS ONE 6, e325 (2008).Article 
CAS 

Google Scholar 
12.Huey, R. B., Losos, J. B. & Moritz, C. Are Lizards Toast?. Science 328, 832–833 (2010).ADS 
CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
13.Sinervo, B. et al. Erosion of lizard diversity by Climate Change and altered thermal niches. Science 328, 894–899 (2010).ADS 
CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
14.Glick, P., Stein, B. A. & Edelson, N. A. Scanning the conservation horizon: A guide to climate change vulnerability assessment (National Wildlife Federation, 2011).
Google Scholar 
15.Parmesan, C. Climate and species’ range. Nature 382, 765–766 (1996).ADS 
CAS 
Article 

Google Scholar 
16.Parmesan, C. et al. Poleward shifts in geographical ranges of butterfly species associated with regional warming. Nature 399, 579–583 (1999).ADS 
CAS 
Article 

Google Scholar 
17.Kelly, A. E. & Goulden, M. L. Rapid shifts in plant distribution with recent climate change. Proc. Natl. Acad. Sci. USA 105, 11823–11826 (2008).ADS 
CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
18.Zuckerberg, B., Woods, A. M. & Porter, W. F. Poleward shifts in breeding bird distributions in New York State. Glob. Change Biol. 15, 1866–1883 (2009).ADS 
Article 

Google Scholar 
19.Sodhi, N. S. & Ehrlich, P. R. Conservation Biology for all (Oxford University Press, 2010).Book 

Google Scholar 
20.Porter, W. P. & Gates, D. M. Thermodynamic equilibria of animals with environment. Ecol. Monogr. 39, 227–244 (1969).Article 

Google Scholar 
21.Avery, R. A. Field studies of body temperatures and thermoregulation. In Biology of the Reptilia (eds Gans, C. & Pough, F. H.) 93–166 (Academic Press, New York, 1982).
Google Scholar 
22.Angilletta, M. J. Thermal Adaptation: A Theoretical and Empirical Synthesis (Oxford University Press, 2009).Book 

Google Scholar 
23.Huey, R. B. Temperature, physiology, and the ecology of reptiles. In Biology of the Reptilia (eds Gans, C. & Pough, F. H.) 25–91 (Academic Press, New York, 1982).
Google Scholar 
24.Angilletta, M. J. Estimating and comparing thermal performance curves. J. Therm. Biol. 31, 541–545 (2006).Article 

Google Scholar 
25.Shine, R. Incubation regimes of cold-climate reptiles: the thermal consequences of nest-site choice, viviparity and maternal basking. Biol. J. Linn. Soc. 83, 145–155 (2004).Article 

Google Scholar 
26.Shine, R. Life-history evolution in Reptiles. Ann. Rev. Ecol. Evol. Syst. 36, 23–46 (2005).Article 

Google Scholar 
27.Huey, R. B. & Stevenson, R. D. Integrating thermal physiology and ecology of ectotherms: A discussion of approaches. Am. Zool. 19, 357–366 (1979).Article 

Google Scholar 
28.Bennett, A. F. The thermal dependence of lizard behaviour. Anim. Behav. 28, 752–762 (1980).Article 

Google Scholar 
29.Christian, K. A. & Tracy, C. R. The effect of the thermal environment on the ability of hatchling Galapagos land iguanas to avoid predation during dispersal. Oecologia 49, 218–223 (1981).ADS 
PubMed 
Article 
PubMed Central 

Google Scholar 
30.Snell, H. L., Jennings, R. D., Snell, H. M. & Harcourt, S. Intrapopulation variation in predator-avoidance performance of Galápagos lava lizards: The interaction of sexual and natural selection. Evol. Ecol. 2, 353–369 (1988).Article 

Google Scholar 
31.Robson, M. A. & Miles, D. B. Locomotor performance and dominance in male Tree Lizards, Urosaurus ornatus. Funct. Ecol. 14, 338–344 (2000).Article 

Google Scholar 
32.Perry, G., LeVering, K., Girard, I. & Garland, T. Locomotor performance and social dominance in male Anolis cristatellus. Anim. Behav. 67, 37–47 (2004).Article 

Google Scholar 
33.Cowles, R. B. & Bogert, C. M. A preliminary study of the thermal requirements of desert reptiles. Bull. Am. Mus. Nat. Hist. 83, 265–296 (1944).
Google Scholar 
34.Bartholomew, G. A. Physiological control of body temperature. In Biology of the Reptilia (eds Gans, C. & Pough, F. H.) 167–211 (Academic Press, New York, 1982).
Google Scholar 
35.Beaupre, S. J. Effects of geographically variable thermal environment on bioenergetics of mottled rock rattlesnakes. Ecology 76, 1655–1665 (1995).Article 

Google Scholar 
36.Huey, R. B., Hertz, P. E. & Sinervo, B. Behavioral drive versus behavioral inertia in evolution: a null model approach. Am. Nat. 161, 357–366 (2003).PubMed 
Article 
PubMed Central 

Google Scholar 
37.Huey, R. B. Behavioral thermoregulation in lizards: importance of associated costs. Science 184, 1001 (1974).ADS 
Article 

Google Scholar 
38.Hertz, P. E., Huey, R. B. & Stevenson, R. D. Evaluating temperature regulation by field-active ectotherms: The fallacy of the inappropriate question. Am. Nat. 142, 796–818 (1993).CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
39.Vitt, L. & Caldwell, J. Herpetology: An introductory biology of amphibians and reptiles 4th edn. (Elsevier, 2014).
Google Scholar 
40.Ortega, Z., Mencía, A. & Pérez-Mellado, V. Sexual differences in behavioral thermoregulation of the lizard Scelarcis perspicillata. J. Therm. Biol. 61, 44–49 (2016).PubMed 
Article 
PubMed Central 

Google Scholar 
41.Rodríguez-Serrano, E., Navas, C. A. & Bozinovic, F. The comparative field body temperature among Liolaemus lizards: Testing the static and the labile hypotheses. J. Therm. Biol. 34, 306–309 (2009).Article 

Google Scholar 
42.Telemeco, R. S., Radder, R. S., Baird, T. A. & Shine, R. Thermal effects on reptile reproduction: Adaptation and phenotypic plasticity in a montane lizard. Biol. J. Linn. Soc. 100, 642–655 (2010).Article 

Google Scholar 
43.Labra, A., Pienaar, J. & Hansen, T. F. Evolution of thermal physiology in Liolaemus Lizards: Adaptation, phylogenetic inertia, and niche tracking. Am. Nat. 174, 204–220 (2009).PubMed 
PubMed Central 
Article 

Google Scholar 
44.Huey, R. B. et al. Why tropical forest lizards are vulnerable to climate warming. Proc. Biol. Sci. 276, 1939–1948 (2009).PubMed 
PubMed Central 

Google Scholar 
45.Deutsch, C. A. et al. Impacts of climate warming on terrestrial ectotherms across latitude. Proc. Natl. Acad. Sci. USA 105, 6668–6672 (2008).ADS 
CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
46.Root, T. L. et al. Fingerprints of global warming on wild animals and plants. Nature 421, 57–60 (2003).ADS 
CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
47.Bestion, E., Teyssier, A., Richard, M., Clobert, J. & Cote, J. Live fast, die young: Experimental evidence of population extinction risk due to climate change. PLoS ONE 13, e1002281 (2015).Article 
CAS 

Google Scholar 
48.Zhang, L., Yang, F. & Zhu, W.-L. Evidence for the ‘rate-of-living’ hypothesis between mammals and lizards, but not in birds, with field metabolic rate. Comp. Biochem. Physiol. Part A 253, 110867 (2021).CAS 
Article 

Google Scholar 
49.Dillon, M. E., Wang, G. & Huey, R. B. Global metabolic impacts of recent climate warming. Nature 467, 704–706 (2010).ADS 
CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
50.Sinervo, B. et al. Climate change, thermal niches, extinction risk and maternal-effect rescue of toad-headed lizards, Phrynocephalus, in thermal extremes of the Arabian Peninsula to the Qinghai—Tibetan Plateau. Integr. Zool. 13, 450–470 (2018).PubMed 
Article 
PubMed Central 

Google Scholar 
51.Ibarguengoytía, N. R. et al. Looking at the past to infer into the future: Thermal traits track environmental change in Liolaemidae. Evolution https://doi.org/10.1111/evo.14246 (2021)Article 
PubMed 
PubMed Central 

Google Scholar 
52.Beniston, M. Climate change in mountain regions: A review of possible impacts. Clim. Change 59, 5–31 (2003).Article 

Google Scholar 
53.Thuiller, W. Climate change and the ecologist. Nature 448, 550–552 (2007).ADS 
CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
54.Martínez Carretero, E. La Puna argentina: Delimitación general y división en distritos florísticos. Bol. Soc. Argent. Bot. 31, 27–40 (1995).
Google Scholar 
55.Esquerré, D., Brennan, I. G., Catullo, R. A., Torres-Pérez, F. & Keogh, J. S. How mountains shape biodiversity: The role of the Andes in biogeography, diversification, and reproductive biology in South America’s most species-rich lizard radiation (Squamata: Liolaemidae). Evolution 73, 214–230 (2019).PubMed 
Article 
PubMed Central 

Google Scholar 
56.Abdala, C. S., Laspiur, A. & Langstroth, R. Las especies del género Liolaemus (Liolaemidae). Lista de taxones y comentarios sobre los cambios taxonómicos más recientes. Cuad. Herp. 35, 193–223 (2021).
Google Scholar 
57.Abdala, C. S. et al. Unravelling interspecific relationships among highland lizards: First phylogenetic hypothesis using total evidence of the Liolaemus montanus group (Iguania: Liolaemidae). Zool. J. Linn. Soc. 189, 349–377 (2020).Article 

Google Scholar 
58.Cabrera, M. R. & Monguillot, J. C. A new Andean species of Liolaemus of the darwinii complex (Reptilia: Iguanidae). Zootaxa 1106, 35–43 (2006).Article 

Google Scholar 
59.Abdala, C. S. et al. Categorización del estado de conservación de las lagartijas y anfisbenas de la República Argentina. Cuad. Herp. 26, 215–248 (2012).
Google Scholar 
60.Avila, L. J. Liolaemus montanezi. The IUCN Red List of Threatened Species 2016: e.T56077261A56077269. https://doi.org/10.2305/IUCN.UK.2016-1.RLTS.T56077261A56077269.en (2016).61.Barros, V. R. et al. Climate change in Argentina: trends, projections, impacts and adaptation. Wiley Interdiscip. Rev. Clim. Change 6, 151–169 (2015).Article 

Google Scholar 
62.IPCC. Climate change 2014: synthesis report. Contribution of Working Groups I, II and III to the Fifh Assessment Report of the Intergovernmental Panel on Climate Change. (IPCC, Geneva, 2014).63.Bury, R. B. Natural history, field ecology, conservation biology and wildlife management: time to connect the dots. Herp. Con. Biol. 1, 56–61 (2006).
Google Scholar 
64.Fei, T. et al. A body temperature model for lizards as estimated from the thermal environment. J. Therm. Biol. 37, 56–64 (2012).Article 

Google Scholar 
65.Ortega, Z. et al. Disentangling the role of heat sources on microhabitat selection of two Neotropical lizard species. J. Trop. Ecol. 35, 149–156 (2019).Article 

Google Scholar 
66.Bujes, C. S. & Verrastro, L. Thermal biology of Liolaemus occipitalis (Squamata, Tropiduridae) in the coastal sand dunes of Rio Grande do Sul, Brazil. Braz. J. Biol. 66, 945–954 (2006).CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
67.Almeida-Santos, P., Militão, C. M., Nogueira-Costa, P., Menezes, V. A. & Rocha, C. F. D. Thermal ecology of five remaining populations of an endangered lizard (Liolaemus lutzae) in different restinga habitats in Brazil. J. Coast. Conserv. 19, 335–343 (2015).Article 

Google Scholar 
68.Liz, A. V., Santos, V., Ribeiro, T., Guimarães, M. & Verrastro, L. Are lizards sensitive to anomalous seasonal temperatures? Long-term thermobiological variability in a subtropical species. PLoS ONE 14, e0226399 (2019).Article 
CAS 

Google Scholar 
69.Martori, R., Bignolo, P. & Cardinale, L. Relaciones térmicas en una población de Liolaemus wiegmannii (Iguania: Tropiduridae). Rev. Esp. Herpetol. 12, 19–26 (1998).
Google Scholar 
70.Martori, R., Aun, L. & Orlandini, S. Relaciones térmicas temporales en una población de Liolaemus koslowskyi. Cuad. Herp. 16, 33–45 (2002).
Google Scholar 
71.Cánovas, M. G., Villavicencio, H. J. & Acosta, J. C. Liolaemus olongasta (NCN) Body temperature. Herp. Rev. 37, 87–88 (2006).
Google Scholar 
72.Villavicencio, H., Acosta, J., Cánovas, M. & Marinero, J. Thermal ecology of a population of the lizard, Liolaemus pseudoanomalus in western Argentina. Amphibia-Reptilia 28, 163–165 (2007).Article 

Google Scholar 
73.Ibargüengoytía, N. R. et al. Thermal biology of the southernmost lizards in the world: Liolaemus sarmientoi and Liolaemus magellanicus from Patagonia, Argentina. J. Therm. Biol. 35, 21–27 (2010).Article 

Google Scholar 
74.Castillo, G. N., Villavicencio, H. J., Acosta, J. C. & Marinero, J. A. Temperatura corporal de campo y actividad temporal de las lagartijas Liolaemus vallecurensis y Liolaemus ruibali en clima riguroso de los Andes centrales de Argentina. Multequina 24, 19–31 (2015).
Google Scholar 
75.Laspiur, A., Villavicencio, H. J. & Acosta, J. C. Liolaemus chacoensis (NCN). Body temperature. Herp. Rev. 38, 458–459 (2007).
Google Scholar 
76.Salva, A. G., Robles, C. I. & Tulli, M. J. Thermal biology of Liolaemus scapularis (Iguania:Liolaemidae) from argentinian northwest. J. Therm. Biol 98, 102924 (2021).PubMed 
Article 

Google Scholar 
77.Mesinger, F., Jovic, D., Chou, S. C., Gomes, J. L. & Bustamante, J. F. Wind forecast around the Andes using the sloping discretization of the eta coordinate. in Proceedings of the 8th International Conference on Southern Hemisphere Meteorology and Oceanography 1837–1848 (INPE, 2006).78.Sannolo, M. & Carretero, M. A. Dehydration constrains thermoregulation and space use in lizards. PLoS ONE 14, e0220384 (2019).CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
79.Nicholson, K. L. et al. The influence of temperatura and humidity on activity patterns of the lizards Anolis stratulus and Ameiva exsul in the British Virgin Islands. Caribb. J. Sci. 41, 870–873 (2005).
Google Scholar 
80.Adolph, A. S. & Porter, W. P. Temperature, activity, and lizard life histories. Am. Nat. 142, 273–295 (1993).CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
81.Bakken, G. S., Santee, W. R. & Erskine, D. Operative and standard operative temperature: Tools for thermal energetics studies. Am. Zool. 25, 933–943 (1985).Article 

Google Scholar 
82.Black, I. R. G., Berman, J. M., Cadena, V. C. & Tattersall, G. J. Behavioral thermoregulation in lizards. Strategies for achieving preferred temperature. In Behavior of lizards. Evolutionary and mechanistic perspectives (eds Bels, V. L. & Russell, A. P.) 13–46 (CRC Press, Florida, 2019).
Google Scholar 
83.Pirtle, E. I., Tracy, C. R. & Kearney, M. R. Hydroregulation. A neglected behavioral response of lizards to climate change? In Behavior of Lizards. Evolutionary and mechanistic perspectives (eds Bels, V. L. & Russell, A. P.) 343–374 (CRC Press, Florida, 2019).
Google Scholar 
84.Medina, M. et al. Thermal biology of genus Liolaemus: A phylogenetic approach reveals advantages of the genus to survive climate change. J. Therm. Biol. 37, 579–586 (2012).Article 

Google Scholar 
85.Blouin-Demers, G. & Weatherhead, P. J. Thermal ecology of black rat snakes (Elaphe obsoleta) in a thermally challenging environment. Ecology 82, 3025–3043 (2001).Article 

Google Scholar 
86.Cabezas-Cartes, F., Fernández, J. B., Duran, F. & Kubisch, E. L. Potential benefits from global warming to the thermal biology and locomotor performance of an endangered Patagonian lizard. PeerJ 7, e7437 (2019).PubMed 
PubMed Central 
Article 

Google Scholar 
87.Obregón, R. L., Scolaro, J. A., Ibargüengoytía, N. R. & Medina, M. Thermal biology and locomotor performance in Phymaturus calcogaster: are Patagonian lizards vulnerable to climate change? Integr. Zool. 16, 53–66 (2021).PubMed 
Article 
PubMed Central 

Google Scholar 
88.Litmer, A. R. & Murray, C. M. Critical thermal tolerance of invasion: Comparative niche breadth of two invasive lizards. J. Therm. Biol. 86, 102432 (2019).PubMed 
Article 
PubMed Central 

Google Scholar 
89.Bels, V. L. & Russell, A. P. Behavior of lizards. Evoutionary and mechanistic perspectives (CRC Press, Florida, 2019).Book 

Google Scholar 
90.Panda, B. B., Achary, V. M., Mahanty, S. & Panda, K. K. Plant adaptation to abiotic and genotoxic stress: Relevance to climate change and evolution. In Climate Change and plant abiotic stress tolerance (eds Tuteja, N. & Gill, S. S.) 251–294 (Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim, 2014).
Google Scholar 
91.Kiesling, R. Flora de San Juan, República Argentina Vol. 1 (Vázquez Mazzini, Buenos Aires, 1994).
Google Scholar 
92.Köeppen, V. P. Climatología. Con un estudio de los climas de la tierra (Fondo de Cultura Económica, Pánuco, México, DF, 1948).
Google Scholar 
93.Bakken, G. S. Measurements and application of operative and standard operative temperatures in ecology. Am. Zool. 32, 194–216 (1992).Article 

Google Scholar 
94.Cecchetto, N. R., Medina, S. M., Taussig, S. & Ibargüengoytía, N. R. The lizard abides: cold hardiness and winter refuges of Liolaemus pictus argentinus in Patagonia, Argentina. Can. J. Zool. 97, 773–782 (2019).Article 

Google Scholar 
95.Cecchetto, N. R., Medina, S. M. & Ibargüengoytía, N. R. Running performance with emphasis on low temperatures in a Patagonian lizard, Liolaemus lineomaculatus. Sci. Rep. 10, 14732 (2020).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
96.Mahoney, J. J. & Hutchison, V. H. Photoperiod acclimation and 24-hour variations in the critical thermal maxima of a tropical and a temperate frog. Oecologia 2, 143–161 (1969).ADS 
CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
97.Christian, K. A. & Weavers, B. W. Thermoregulation of monitor lizards in Australia: An evaluation of methods in thermal biology. Ecol. Monogr. 66, 139–157 (1996).Article 

Google Scholar 
98.Camacho, A. et al. Measuring behavioral thermal tolerance to address hot topics in ecology, evolution, and conservation. J. Therm. Biol. 73, 71–79 (2018).PubMed 
Article 
PubMed Central 

Google Scholar 
99.Clusella-Trullas, S. & Chown, S. L. Lizard thermal trait variation at multiple scales: a review. J. Comp. Physiol. B 184, 5–21 (2014).PubMed 
Article 
PubMed Central 

Google Scholar 
100.Sunday, J. M. et al. Thermal-safety margins and the necessity of thermoregulatory behavior across latitude and elevation. Proc. Natl. Acad. Sci. USA 111, 5610–5615 (2014).ADS 
CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
101.Sokal, R. R. & Rohlf, F. J. Biometry: The Principles and practice of statistics in biological research 3rd edn. (Freeman W.H, 1995).MATH 

Google Scholar 
102.Kovach, W. Oriana ver. 4.0. Software. (Kovach Computing Services, 2001).103.Fitzgerald, L. A., Cruz, F. B. & Perotti, G. Phenology of a lizard assemblage in the dry Chaco of Argentina. J. Herpetol. 33, 526–535 (1999).Article 

Google Scholar 
104.Beasley, T. M. & Schumacker, R. E. Multiple regression approach to analyzing contingency tables: Post hoc and planned comparison procedures. J. Exp. Educ. 64, 79–93 (1995).Article 

Google Scholar 
105.García Pérez, M. A. & Núñez Antón, V. Cellwise residual analysis in two-way contingency tables. Educ. Psychol. Meas. 65, 825–839 (2003).MathSciNet 
Article 

Google Scholar 
106.Peig, J. & Green, A. J. New perspectives for estimating body condition from mass/length data: the scaled mass index as an alternative method. Oikos 118, 1883–1891 (2009).Article 

Google Scholar 
107.Bohonak, A. J. & van Der Linde, K. RMA for JAVA Software for Reduced Major Axis regression. ver. 1.21. (2004).108.Baty, F. et al. A toolbox for nonlinear regression in R: The Package nlstools. J. Stat. Softw. 5, 1–21 (2015).
Google Scholar 
109.R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing (2019).110.Thuiller, W., Georges, D., Engler, R. & Breiner, F. biomod2: Ensemble Platform for Species Distribution Modeling. R package ver. 3.4.6. (2020).111.Hijmans, R. J., Phillips, S., Leathwick, J. & Elith, J. dismo: Species Distribution Modeling. R package ver. 1.1-4. (2017).112.Hijmans, R. J. raster: Geographic Data Analysis and Modeling. R package ver. 3.4-5. (2020). More

Fieldwork and rock sample analysisThe primary objective of our fieldwork was to collect sedimentological data that would allow us to interpret the processes responsible for the deposition of the beds of the Greater Phyllopod Bed. These parameters could then be incorporated into our experimental design and recreation of Burgess Shale-type flows. To understand the complex sedimentary deposits of the Burgess Shale Formation, we targeted individual beds (Fig. 3, Supplementary Figs. 2–5) that were logged at outcrop for informative mm-scale and cm-scale sedimentary structures. Grain size analysis was conducted in the field using a grain-size comparator and hand-lens and during petrographic analysis. The Greater Phyllopod Bed has been logged in considerable detail in the field20,33, and so logs produced from our work can be used to compare to previous studies. Detailed descriptions of the intervals sampled included color, bounding surfaces, micro-sedimentary structures, grain size, and textures. Larger-scale field mapping and analysis of sedimentary architecture were not undertaken and so we were not attempting to answers questions on the relationship of the Cathedral Escarpment to the fossil-bearing deposits or the precise provenance of the organisms.We collected whole-rock samples from the Greater Phyllopod Bed of the Walcott Quarry at stratigraphic heights of 111.6, 136, 149.95, 184.83, and 226.68 cm (labeled Bed A to E, respectively) above the top of the Wash Limestone Member. All sedimentological samples for this study were collected in situ from this location under the Parks Canada collection and research permit (YNP-2015-19297). The permit for our fieldwork allowed us to collect and sample sedimentological material exclusively. These were subsequently sampled for laboratory analysis and thin-section preparation.Petrographic analysis was performed on all samples using a Leica DM750P microscope. Each thin section was scanned with an Epson scanner to observe details of the millimeter-scale structures and textures (Fig. 3, Supplementary Figs. 2–5). Plain and cross-polarized light micrographs were taken of areas of particular sedimentological interest from each thin section and documented along with the petrological analysis. These samples were processed for further geochemical and elemental analysis.Sample analysisX-Ray Diffraction (XRD) was used to characterize the mineralogical content of the matrix of Bed A (111.6 cm above the top of the Wash Limestone Member) from the Walcott Quarry. For whole-rock bulk powder analyses, the sample was ground into a powder, and XRD was conducted using a PANalytical X’Pert3 diffractometer. For clay analysis, we applied the fractions to orientated glass slides. Organics were removed from each sample by H2O2 treatment before disaggregating the material using ultrasonic vibration. The suspended material was decanted from the ultrasonic bath in centrifuge bottles, which were topped up with deionized water so that each bottle weighed within the same gram. The bottles were placed in the centrifuge for two treatments, first at 1000 rpm for 4 min, and then again at 4000 rpm for 20 min. After the first treatment, the supernatant was transferred to new centrifuge bottles. The three lightest bottles were topped up with deionized water in order to reach the weight of the heaviest. The resultant concentrated sample yield ( More

Selection of materials for joint repair of chromium-contaminated soilTable 1 shows the results of the orthogonal test. Range analysis was performed according to the results of Table 1. The range-analysis results are shown in Table 2.Table 1 Orthogonal design scheme and results.Full size tableTable 2 Orthogonal test results range analysis calculation table.Full size tableTable 2 shows that, from the perspective of unconfined compressive strength, the primary and secondary order of the 28 day strength, factors affecting the combined repair of chromium-contaminated soil were cement content → fly-ash synthetic zeolite content → CaS5 content. The best test ratio was: CaS5 content 3 times, synthetic zeolite content 15%, and cement content 20%. The unconfined compressive strength of the contaminated soil after remediation increased with the increase in cement content, but the relationship between the content of CaS5 and synthetic zeolite, and the unconfined compressive strength of the specimen was not very obvious. From the perspective of toxicity leaching, the primary and secondary order of factors affecting the total chromium leaching concentration of the combined remediation of chromium-contaminated soil were cement content → fly-ash synthetic zeolite content → CaS5 content. The primary and secondary order of factors affecting the leaching concentration of Cr(VI) in the combined remediation of contaminated soil were CaS5 content → cement content → fly-ash synthetic zeolite content. The best test ratios of the total chromium and Cr(VI) toxicity leaching test were: CaS5 content is 4 times, synthetic zeolite content 15%, and cement content 20%. Total chromium and Cr(VI) leaching concentration of the chromium-contaminated soil after joint remediation was negatively correlated with the content of CaS5, synthetic zeolite, and cement content. The change of total chromium leaching concentration was most significantly affected by cement content and synthetic zeolite. Second, the change of Cr(VI) leaching concentration was most significantly affected by CaS5 content. From the perspective of leaching concentration, when reducing agent CaS5, adsorbent synthetic zeolite, and curing agent cement were all at maximum, the leaching effect of total chromium and Cr(VI) was best. However, considering the actual engineering cost and dosage of the preparation should be reduced as much as possible for meeting the requirements. Therefore, comprehensive balance analysis determined the optimal ratio for joint repair of chromium-contaminated soil to be 3 times the dosage of CaS5, 15% synthetic zeolite, and cement amount 20%.Strength change of combined repair of chromium-contaminated soil under action of dry–wet cycleThe test compared the variation of unconfined compressive strength with the number of dry and wet cycles under different conditions of chromium content, combined to repair standard specimens of chromium-contaminated soil, and test results are shown in Fig. 1.Figure 1The relationship between unconfined compressive strength and the number of dry wet cycles.Full size imageFigure 1 shows that, in the beginning, the unconfined compressive strength of the combined repair of chromium-contaminated soil increased with the increase in the number of wet and dry cycles. After reaching the maximal value, it gradually decreased as the number of dry–wet cycles continued to increase. In the initial stage of the dry–wet cycle, the unconfined compressive strength of the combined repair of chromium-contaminated soil increased to varying degrees. For 1000 and 3000 mg/kg of chromium-contaminated soil, the peak of the unconfined compressive strength appeared at 2 times during the dry–wet cycle, and the peak of the unconfined compressive strength of 5000 mg/kg chromium-contaminated soil appeared at 4 dry–wet cycles. After that, unconfined compressive strength gradually decreased with the progress of dry–wet cycles, and the decrease rate became slower. From strength-loss analysis, the higher the chromium content was, the greater the change in strength loss. After 16 wet and dry cycles, the strength-loss rates of 1000, 3000, and 5000 mg/kg chromium-contaminated soil were 17.95%, 22.27%, and 28.73%, respectively, and strength loss was within 30%, showing better water stability21,22.From analysis of the strength-change process, after 28 days of curing for the joint repair of chromium-contaminated soil, the physical and chemical interaction between cement hydrate and soil in the repair preparation was still occurring, as was the strength increase and dry–wet cycle caused by its hydration products. The weakening effect on strength is a dynamic equilibrium process of mutual decline and growth, and the equilibrium state of the two reaction degrees directly affected the strength of solidified chromium-contaminated soil23. In the initial stage of the dry–wet cycle, the strength increase caused by the interaction between remediation agent and chromium-contaminated soil continued. At that time, the destructive effect of the dry–wet cycle on the joint repair of chromium-contaminated soil was not significant in comparison. As the number of dry–wet cycles increased, hydration products formed and became stable. Dry shrinkage and wet expansion cause internal stress in the joint repair of chromium-contaminated soil, and the soil has cracks due to internal stress changes. A dry–wet cycle has a relatively destructive effect that is gradually noticeable and resulting in a decrease in strength. After many instances of drying and wetting, the strength of repairing chromium-contaminated soil was decreased and stabilized.Figure 1 also shows that, compared with low-content chromium-contaminated soil, the high-content chromium-contaminated-soil solidified body strength peak appeared later, and the peak value was low. This is because the higher the chromium ion content was, the more serious the delay of the hydration reaction of the repair agent was, and the more obvious the weakening effect on the strength of the cured body was, which is not conducive to strength growth. The weakening effect of the dry–wet cycle on strength continued to exist, which led to the repaired contaminated soil with a high content of chromium having lower strength.Toxic-leaching changes of combined remediation of chromium-contaminated soil under dry–wet cycleThe experiment compared the variation of hexavalent chromium and total chromium leaching concentration with the number of dry–wet cycles in standard specimens of the combined repair of chromium-contaminated soil under different chromium-content conditions of the contaminated soil. Test results are shown in Fig. 2.Figure 2Effect of drying–wetting cycle timeson leaching concentration of Cr.Full size imageFigure 2 shows that the leaching concentration of Cr(VI) and total chromium decreased in the initial stage of the dry–wet cycle of the remediation of chromium-contaminated soil. After that, as the number of dry–wet cycles increased, leaching concentration also increased, but the content was low (1000 mg/kg). The medium content (3000 mg/kg) of chromium-contaminated soil Cr(VI) and total chromium leaching concentration fluctuated slightly, and the change was relatively stable, while the high content of chromium-contaminated soil (5000 mg/kg) Cr(VI) leaching the concentration fluctuated greatly, and total chromium increased significantly. Compared with the low-content chromium-contaminated soil, the leaching concentration of the solidified body of high-content chromium-contaminated soil was higher.In the beginning of the dry–wet cycle, the physical and chemical interaction between the cement hydrate and the soil in the repair preparation was still happening. The fly-ash synthetic zeolite had the adsorption effect of metal chromium ions and hydroxide precipitation in the alkaline environment. The formation of chromium ions could meet the requirements of curing/stabilizing chromium ions, and heavy-metal chromium ions are not easy to leach. With the increase in the number of dry–wet cycles, a series of evolutionary processes occurred, such as the expansion of local microcracks, the increase in macropores, the appearance of internal cracks in the contaminated soil, and the appearance of cracks and peeling phenomena on the outside of the contaminated-soil damage. At this time, the contact area between the heavy-metal ions in the contaminated soil and the external environment, especially water, increased, which reduced the ability of the repair agent to adsorb and wrap chromium ions, so that chromium ions were easily leached. In the leaching test, the use of the acidic leaching solution also destroyed the pH balance of the repaired chromium-contaminated soil, the hydrated gel was dissolved and desorbed, and the heavy metals changed, thereby accelerating the leaching of heavy-metal ions24.From analysis of the leaching law shown by the contaminated soil with different chromium content levels, when chromium content in the contaminated soil was low, the remediation agent could effectively solidify/stabilize most of the chromium ions in the soil Cr(VI) and low total chromium leaching. When the chromium content in the contaminated soil was high, the limited content of the repair agent showed an insufficient solidification/stabilization effect of the heavy-metal chromium ions. Because a higher concentration of chromium ions hindered the formation of hydration products of the repair agent, it weakened the adsorption and binding capacity of the hydrated gel. The heavy-metal chromium ions existed in the pores of the contaminated soil in a free state, making the repair agent solidify the chromium ions, the stabilization effect decreased, and the leaching of Cr(VI) and total chromium increased.Overall, the effect of the dry–wet cycle on the joint repair of chromium-contaminated soil was limited, and the joint repair of chromium-contaminated soil had strong resistance to dry–wet cycles, especially the low- and medium-content chromium-polluted soil.Combined repair of quality loss of chromium-contaminated soil under action of dry–wet cyclesThe cumulative mass-loss rate of the sample was calculated from Formula (1), and the result is shown in Fig. 3. With the increase in the number of wet and dry cycles, the cumulative mass-loss rate of the composite preparation to repair chromium-contaminated soil gradually increased; and the higher the chromium content of the contaminated soil was, the greater the cumulative mass-loss rate was. The cumulative mass-loss rate of 16 wet and dry cycles was less than 1%, which shows that the joint repair of chromium-contaminated soil had strong resistance to dry and wet cycles.Figure 3Change of cumulative mass loss rate during dry wet cycle.Full size imageFigure 4 is a photograph of the appearance change of a solidified 5000 mg/kg chromium-contaminated-soil sample after a dry–wet cycle. The soundness-evaluation results of the sample after each dry–wet cycle are shown in Fig. 5.Figure 4Appearance changes of cured chromium contaminated soil samples with dry and wet cycles at (a) 0 times; (b) 2 times; (c) 4 times; (d) 8 times; and (e) 16 times.Full size imageFigure 5Soundness evaluation results of cured chromium contaminated soil samples.Full size imageFigures 4 and 5 show that, after two dry–wet cycles of the joint repair of chromium-contaminated soil, the appearance of the sample did not significantly change, compared with 0 cycles, the surface changed from smooth to rough. Slight cracks appeared from the fourth cycle. Obvious cracks appeared in the sample at the end of the eighth cycle, and a small part of the sample fell off. The sample began to show obvious cracks from the end of the 15th dry–wet cycle, and large pieces of slack simultaneously appeared. The sample was subjected to 16 wet and dry cycles, and soundness was still not at e–h level, indicating that the joint repair of chromium-contaminated soil had strong resistance to dry and wet cycles.Combined repair of chromium-contaminated-soil microstructure changes under action of dry–wet cyclesAfter the joint repair of chromium-contaminated-soil specimens underwent a certain number of wet and dry cycles, the strength, leaching characteristics, and appearance of the specimens significantly changed. From the microstructure, there had to be corresponding changes. Therefore, scanning electron microscope (SEM) and X-ray diffraction (XRD) were used to further analyze the microstructure changes of specimens with different chromium content levels under the action of different wet and dry cycles, as shown in Figs. 6 and 7.Figure 6SEM images of 5000 mg/kg chromium contaminated soil specimens after different dry wet cycles at (a) 0 times; (b) 2 times; (c) 8 times; and (d) 16 times.Full size imageFigure 7XRD pattern of 5000 mg/kg chromium contaminated soil specimen after different dry wet cycles.Full size imageFigure 6 shows that the combined repair of chromium-contaminated soil after 28 days of curing had many pores in the specimen at 0 dry–wet cycles (standard sample), the physical and chemical interaction between the cement hydrate and the soil in the repair preparation still continued, and there were platelike calcium hydroxide crystals on the surface. After two dry–wet cycles, the contaminated soil was denser, and the overall structure was more complete than that in the samples without dry–wet cycles. The plate-shaped calcium hydroxide crystals were reduced, and a large number of fibrous and flocculent hydrated gels could be seen on the surface of the structure. This shows that the reaction between remediation agent and chromium-contaminated soil continued, which is consistent with the law that strength did not drop but rose during the two dry and wet cycles in the unconfined-compressive-strength test. After the test piece had undergone 8 dry–wet cycles, the surface of the test piece not only had a large increase in pores, but also had local cracks, indicating that the structure of the test piece was damaged under the action of the dry–wet cycle, which is consistent with the unconfined compressive strength found in the experiment, coinciding with a sharp drop. After 16 wet and dry cycles, the surface of the specimen not only showed a large number of pores and cracks, but also had obvious roughness. It showed that the dry–wet cycle effect caused the hydration products and cement materials in the soil to be destroyed and dissolved out, and the coupling and supporting forces between soil particles are weakened, and the strength of the soil is reduced accordingly, which was consistent with the macroscopic test results.Figure 7 shows that the main crystal phases of the chromium-contaminated soil were SiO2 and Al2O3 for the samples that did not undergo a dry–wet cycle. A small number of CSH, CAH, Ca(OH)2, and CaCO3 crystals could also be detected from the diffraction peaks. Cr3+ and Cr6+ formed hydroxide precipitates in a highly alkaline environment and wrapped them on the surface of cement, hindering their contact reaction with water. Compared with 0 cycles, SiO2 and Al2O3 in the second cycle were decreased, while the contents of CSH, CAH, Ca(OH)2, and CaCO3 significantly increased. This is because in the process of dry and wet cycles, the sample is fully exposed to moisture and air, so the hydration, depolymerization-cementation, pozzolanic, and carbonation reactions between composite preparation and chromium-contaminated soil continued. After two dry–wet cycles, more hydration products were generated than in the specimens without dry–wet cycles, which filled the pores between the particles of the solidified body, effectively blocking the permeability of the pores, and making the contaminated soil denser, and more structured and complete. At the same time, the full progress of the hydration reaction also delayed the damage rate of the water body to the soil in the dry–wet cycle, so that the soil could maintain a certain strength in the harsh environment, which is consistent with the above-mentioned growth trend of the soil strength. At the same time, the extension of a large amount of fibrous calcium silicate hydrate greatly increased the internal specific surface area of the soil. Free-state Cr3+ and Cr6+ were adsorbed or produced hydroxide precipitation and filled in the pores of the soil, and free ion concentration was also greatly reduced, which is consistent with the above ion-leaching test results. For the specimens with 8 dry and wet cycles, the content of hydration products such as CAH and CSH was reduced. This is due to a series of evolutionary processes such as the expansion of local microcracks, the increase in macropores, the appearance of internal cracks in the contaminated soil, and the appearance of cracks and peeling on the outside of the contaminated soil. Structural integrity was destroyed, and strength was accordingly reduced. By 16 wet and dry cycles, a large amount of fibrous CSH disappeared, which weakened the cementation between soil particles. At this time, the heavy-metal ions originally wrapped in the contaminated soil solidified the body and the external environment, the contact area with the water was increased, the pH value of the environment was decreased, hydrate CSH was decalcified, and Ca/Si ratio was decreased. This reduced the adsorption capacity of the compound formulation to chromium ions, so that chromium ions were dissolved out of the soil. More

Wildlife Conservation Research Unit, Recanati-Kaplan Centre, Zoology, University of Oxford, Oxford, UKHans Bauer & Claudio Sillero-ZubiriEvolutionary Ecology Group, Biology, University of Antwerp, Antwerp, BelgiumHans BauerLaboratory for Applied Ecology, Natural Resource Conservation, University of Abomey-Calavi, Cotonou, BeninAristide Comlan TehouDepartment of HydroSciences and Environment, University Iba Der Thiam, Thiès, SénégalMallé GueyeDirection de la Faune, de la Chasse et des Parcs et Réserves, Ministère de l’Environnement de la Salubrité Urbaine et du Développement Durable, Niamey, NigerHamissou GarbaDirection de la Faune et des Chasses, Ministère de l’Environnement et du Développement Durable, Ouagadougou, Burkina FasoBenoit DoambaNational Parks Directorate, Ministry of Environment and Sustainable Development, Dakar, SenegalDjibril DiouckThe Born Free Foundation, Horsham, UKClaudio Sillero-Zubiri More

1.Wetzel, R. G. In Limnology 3rd edn (ed. Wetzel, R. G.), Ch. 9, 151–168 (Academic Press, 2001).2.Schindler, D. Warmer climate squeezes aquatic predators out of their preferred habitat. Proc. Natl Acad. Sci. USA 114, 9764–9765 (2017).CAS 
Article 
ADS 

Google Scholar 
3.North, R. P., North, R. L., Livingstone, D. M., Köster, O. & Kipfer, R. Long-term changes in hypoxia and soluble reactive phosphorus in the hypolimnion of a large temperate lake: consequences of a climate regime shift. Glob. Change Biol. 20, 811–823 (2014).Article 
ADS 

Google Scholar 
4.Fernández, J. E., Peeters, F. & Hofmann, H. Importance of the autumn overturn and anoxic conditions in the hypolimnion for the annual methane emissions from a temperate lake. Environ. Sci. Technol. 48, 7297–7304 (2014).Article 
ADS 

Google Scholar 
5.Michalak, A. M. et al. Record-setting algal bloom in Lake Erie caused by agricultural and meteorological trends consistent with expected future conditions. Proc. Natl Acad. Sci. USA 110, 6448–6452 (2013).CAS 
Article 
ADS 

Google Scholar 
6.Schmidtko, S., Stramma, L. & Visbeck, M. Decline in global oceanic oxygen content during the past five decades. Nature 542, 335–339 (2017).CAS 
Article 
ADS 

Google Scholar 
7.Breitburg, D. et al. Declining oxygen in the global ocean and coastal waters. Science 359, (2018).8.Jankowski, J., Livingstone, D. M., Bührer, H., Forster, R. & Niederhauser, P. Consequences of the 2003 European heat wave for lake temperature profiles, thermal stability, and hypolimnetic oxygen depletion: implications for a warmer world. Limnol. Oceanogr. 51, 815–819 (2006).Article 
ADS 

Google Scholar 
9.Yvon-Durocher, G., Jones, J. I., Trimmer, M., Woodward, G. & Montoya, J. M. Warming alters the metabolic balance of ecosystems. Phil. Trans. R. Soc. B 365, 2117–2126 (2010).Article 

Google Scholar 
10.Seki, H., Takahashi, Y., Hara, Y. & Ichimura, S. Dynamics of dissolved oxygen during algal bloom in Lake Kasumigaura, Japan. Water Res. 14, 179–183 (1980).CAS 
Article 

Google Scholar 
11.Jacobson, P. C., Stefan, H. G. & Pereira, D. L. Coldwater fish oxythermal habitat in Minnesota lakes: influence of total phosphorus, July air temperature, and relative depth. Can. J. Fish. Aquat. Sci. 67, 2002–2013 (2010).CAS 
Article 

Google Scholar 
12.Harke, M. J. et al. A review of the global ecology, genomics, and biogeography of the toxic cyanobacterium, Microcystis spp. Harmful Algae 54, 4–20 (2016).Article 

Google Scholar 
13.Vaquer-Sunyer, R. & Duarte, C. M. Thresholds of hypoxia for marine biodiversity. Proc. Natl Acad. Sci. USA 105, 15452–15457 (2008).CAS 
Article 
ADS 

Google Scholar 
14.Woolway, R. I. & Merchant, C. J. Worldwide alteration of lake mixing regimes in response to climate change. Nat. Geosci. 12, 271–276 (2019).CAS 
Article 
ADS 

Google Scholar 
15.Livingstone, D. M. Impact of secular climate change on the thermal structure of a large temperate central European lake. Clim. Change 57, 205–225 (2003).Article 

Google Scholar 
16.Zhang, Y. et al. Dissolved oxygen stratification and response to thermal structure and long-term climate change in a large and deep subtropical reservoir (Lake Qiandaohu, China). Water Res. 75, 249–258 (2015).CAS 
Article 

Google Scholar 
17.Bouffard, D., Ackerman, J. D. & Boegman, L. Factors affecting the development and dynamics of hypoxia in a large shallow stratified lake: hourly to seasonal patterns. Wat. Resour. Res. 49, 2380–2394 (2013).CAS 
Article 
ADS 

Google Scholar 
18.O’Reilly, C. M. et al. Rapid and highly variable warming of lake surface waters around the globe. Geophys. Res. Lett. 42, 10773–10781 (2015).ADS 

Google Scholar 
19.Nürnberg, G. K. Trophic state of clear and colored, soft- and hardwater lakes with special consideration of nutrients, anoxia, phytoplankton and fish. Lake Reserv. Manage. 12, 432–447 (1996).Article 

Google Scholar 
20.Ho, J. C., Michalak, A. M. & Pahlevan, N. Widespread global increase in intense lake phytoplankton blooms since the 1980s. Nature 574, 667–670 (2019).CAS 
Article 
ADS 

Google Scholar 
21.Kosten, S. et al. Warmer climates boost cyanobacterial dominance in shallow lakes. Glob. Change Biol. 18, 118–126 (2012).Article 
ADS 

Google Scholar 
22.Müller, B., Bryant, L. D., Matzinger, A. & Wüest, A. Hypolimnetic oxygen depletion in eutrophic lakes. Environ. Sci. Technol. 46, 9964–9971 (2012).PubMed 

Google Scholar 
23.Winslow, L. A., Leach, T. A. & Rose, K. C. Global lake response to the recent warming hiatus. Environ. Res. Lett. 13, 054005 (2018).Article 
ADS 

Google Scholar 
24.Livingstone, D. M. An example of the simultaneous occurrence of climate-driven “sawtooth” deep-water warming/cooling episodes in several Swiss lakes. Verh. Int. Ver. Limnol. 26, 822–828 (1997).
Google Scholar 
25.Williamson, C. E. et al. Ecological consequences of long-term browning in lakes. Sci. Rep. 5, (2015).26.Rose, K. C., Winslow, L. A., Read, J. S. & Hansen, G. J. A. Climate-induced warming of lakes can be either amplified or suppressed by trends in water clarity. Limnol. Oceanogr. Lett. 1, 44–53 (2016).Article 

Google Scholar 
27.Woolway, R. I. et al. Northern hemisphere atmospheric stilling accelerates lake thermal responses to a warming world. Geophys. Res. Lett. 46, 11983–11992 (2019).Article 
ADS 

Google Scholar 
28.Carpenter, S. R., Stanley, E. H. & Vander Zanden, M. J. State of the world’s freshwater ecosystems: physical, chemical, and biological changes. Annu. Rev. Environ. Resour. 36, 75–99 (2011). Article 

Google Scholar 
29.R Core Team. R: A Language and Environment for Statistical Computing. http://www.R-project.org/ (R Foundation for Statistical Computing, Vienna, 2017).30.Borchers, H. W. pracma: Practical Numerical Math Functions. R package version 2.1.5 https://CRAN.R-project.org/package=pracma (2018).31.Winslow, L. A. et al. rLakeAnalyzer: Lake Physics Tools. R package version 1.11.4. https://CRAN.R-project.org/package=rLakeAnalyzer (2017).32.Winslow, L. A. et al. LakeMetabolizer: an R package for estimating lake metabolism from free-water oxygen using diverse statistical models. Inland Waters 6, 622–636 (2016).CAS 
Article 

Google Scholar 
33.Carslaw, D. C. & Ropkins, K. Openair – an R package for air quality data analysis. Environ. Model. Softw. 27-28, 52–61 (2012).Article 

Google Scholar 
34.Moran, P. A. P. The interpretation of statistical maps. J. R. Stat. Soc. B 10, 243–251 (1948).MathSciNet 
MATH 

Google Scholar 
35.Kalogirou, S. lctools: Local Correlation, Spatial Inequalities, Geographically Weighted Regression and Other Tools. R package version 0.2-7. https://CRAN.R-project.org/package=lctools (2019).36.Copernicus Climate Change Service (C3S). ERA5: Climate Data Store (CDS) https://cds.climate.copernicus.eu/cdsapp#!/home (accessed 1 October 2019).37.Gelman, G. & Hill, J. Data Analysis Using Regression and Multilevel/Hierarchical Models (Cambridge Univ. Press, 2007).38.Quinn, G. P. & Keough, M. J. Experimental Design and Data Analysis for Biologists (Cambridge Univ. Press, 2002).39.Lumley, T. leaps: Regression Subset Selection. R package version 3.1. https://CRAN.R-project.org/package=leaps (2020).40.Wood, S. N. Generalized Additive Models: An Introduction with R 2nd edn (CRC Press, 2017).Book 

Google Scholar 
41.Wood, S. & Scheipl, F. gamm4: Generalized Additive Mixed Models using ‘mgcv’ and ‘lme4’. R package version 0.2-5. https://CRAN.R-project.org/package=gamm4 (2017).42.Pinheiro, J. C. & Bates, D. M. Mixed Effects Models in S and S-Plus (Springer, 2000).43.Burnham, K. P., Anderson, D. R. & Huyvaert, K. P. AIC model selection and multimodel inference in behavioral ecology: some background, observations, and comparisons. Behav. Ecol. Sociobiol. 65, 23–35 (2011).Article 

Google Scholar 
44.Hosmer, D. W. & Lemeshow, S. Applied Logistic Regression 2nd edn (John Wiley and Sons, Inc., 2000).45.Homer, C. G. et al. Completion of the 2011 National Land Cover Database for the conterminous United States – Representing a decade of land cover change information. Photogramm. Eng. Remote Sensing 81, 345–354 (2015).
Google Scholar 
46.Lele, S. R., Keim, J. L. & Solymos, P. ResourceSelection: Resource Selection (Probability) Functions for Use-Availability Data. R package version 0.3-2. https://CRAN.R-project.org/package=ResourceSelection (2017).47.Cutler, D. R. et al. Random forests for classification in ecology. Ecology 88, 2783–2792 (2007).Article 

Google Scholar 
48.Liaw, A. & Wiener, M. Classification and regression by randomForest. R News 2, 18–22 (2002).
Google Scholar 
49.Messager, M. L., Lehner, B., Grill, G., Nedeva, I. & Schmitt, O. Estimating the volume and age of water stored in global lakes using a geo-statistical approach. Nat. Commun. 7, 13603 (2016).CAS 
Article 
ADS 

Google Scholar 
50.Stetler, J. T., Jane, S. F., Mincer, J. L., Sanders, M. N. & Rose, K. C. Long-term lake dissolved oxygen and temperature data, 1941–2018 ver 2. Environmental Data Initiative https://doi.org/10.6073/pasta/841f0472e19853b0676729221aedfb56 (2021).51.Adrian, R., Jane, S. F., & Rose, K. C. Widespread deoxygenation of temperate lakes – Müggelsee data. IGB Leibniz-Institute of Freshwater Ecology and Inland Fisheries dataset. https://doi.org/10.18728/568.0 (2021).52.Jenny, J.-P. Time series dataset of dissolved oxygen, water temperature and Secchi depths profiles in Lakes Annecy and Geneva. Portail Data INRAE V1, https://doi.org/10.15454/BUJUSX (2021). More

A review of existing despiking proceduresAmong despiking algorithms for raw, high-frequency, EC data, a popular approach was developed by Vickers and Mahrt6 (hereinafter VM97). The method consists in estimating the sample mean and standard deviation in overlapping temporal windows whose width in time is 5 min. The temporal window slides point by point, and any data point whose value exceeds (pm 3.5 sigma) (sample standard deviation) is flagged as a spike. The method is highly sensitive to the masking effect (where less extreme spikes go undetected because of the existence of the most extreme spikes), a reason for which the procedure is iterated increasing by 0.1 the threshold value at each pass, until no more spikes are detected.A revised version of the VM97 procedure was proposed by Metzger et al.14 (hereinafter M12), who suggested replacing the mean and standard deviation by more robust estimates, such as the median and the median absolute deviation (MAD), respectively. The authors found that this method reliably removed spikes that were not detected by VM97, showing a superior performance.To reduce the high-computational burden attributable to the windowed computations prescribed by the VM97 algorithm, Mauder et al.7 (hereinafter M13) proposed to estimate median and MAD over the whole flux averaging period (usually 30 or 60 min). M13 suggested to consider as spike those observations exceeding (pm 7cdot)MAD. Such an approach was selected as candidate method in the data processing scheme at the ICOS ecosystem stations15.Starkenburg et al9 recommended the approach developed by Brock16 (hereinafter BR86) as the best method for despiking EC data. This algorithm is currently implemented in the processing pipeline adopted by the National Ecological Observatory Network (NEON, https://www.neonscience.org). It is based on a two-stage procedure, where the first step consists in extracting the signal by means of a rolling third-order median filter which replaces the center value in the window with the median value of all the points within the window; the second step aims at identifying spikes by analyzing the histogram of the differences between the raw signal and the median filtered signal. Specifically, the differences are initially binned into 25 classes. Then, the first bins with zero counts on either side of the histogram are identified and points in the original signal that exceed the empty bins are flagged as spikes. If no bin with zero counts is found, then the number of bins is doubled (for example from 25 to 51, with one bin added ensuring to retain an odd number because the mean of differences, which is expected to be close to zero, should fall into the central bin of the histogram). The procedure is iterated by increasing the number of bins until the bin width is not less than the acquiring instrument resolution.The proposed despiking algorithmFigure 1Flowchart of the proposed despiking algorithm.Full size image
In order to define a modeling framework suitable for the representation of a sequence ((x_t)_{t in Z}) of observed raw EC data indexed by time t and contaminated by spikes, we assume a component model as follows:$$begin{aligned} left. begin{aligned} x_t&= mu _t + v_t + s_t,\ end{aligned}right. end{aligned}$$
(1)
where (mu _t) denotes the low frequency component (signal); (v_t) the deviations from the signal level (residuals) whose variability ((sigma _t^2)) is allowed to change slowly over time; and (s_t) the spike generating mechanism which is zero most of time but occasionally generates large absolute values.To achieve unbiased estimates of both the signal and the scale parameter ((sigma _t)) when data are contaminated by errors, the use of robust estimators is required. One of the most popular measures of robustness of a statistical procedure is the breakdown point, which represents the proportion of outlying data points an estimator can resist before giving a biased result. The maximum breakdown point is 50%, since, if more than half of the observations are contaminated, it is not possible to distinguish between the distribution of good data and the distribution of outlying data. Described in these terms, the arithmetic mean has a breakdown point of 0% (i.e. we can make the mean arbitrarily large just by changing any of the data point), whereas the median has a breakdown point of 50% (i.e. it becomes biased only when 50% or more of the data are large outliers).The proposed despiking procedure (hereinafter RobF) makes use of robust functionals whose breakdown point is 50% and consists in three stages (see Fig. 1). In the first step the signal ((mu _t)) extraction is carried out by means of the repeated median (RM) regression technique10,17. The second step involves the estimation of the time-varying scale parameter (sigma _t) by means of the (Q_n) estimator12. A detailed description of the robust functionals will be provided in the following sections. Spikes are detected in the third step, through the examination of outlier scores calculated as:$$begin{aligned} z_t=frac{x_t-mu _t}{sigma _t}. end{aligned}$$
(2)
Any values of (|z_t|) exceeding a pre-fixed threshold value ((z_{th})) is considered as spike. The choice of the threshold value should be based on the outlier scores data distribution which can vary across time. In this work (z_{th}) was set equal to 5 which means that for Normal- and Laplace-distributed data there is a 1 in 3.5 million and 1 in 300 chance, respectively, that an anomalous value is the result of a statistical fluctuation over the spectrum of plausible values. Once detected, spikes are removed and replaced by (mu _t) estimates obtained by the RM filter.Repeated median filterThe idea underlying moving time window based approaches is that of approximating the signal underlying observed data by means of local estimates that approximate the level of data in the center of the window.To this end, we fit a local linear trend11 of the form$$begin{aligned} mu _{t+i}=mu _t+ibeta _t, quad i=-k,ldots ,k, quad mathrm {to} quad {x_{t-k},ldots ,x_{t+k}}, end{aligned}$$
(3)
where k is the parameter defining the time window of length (n=2k+1), whereas (mu _t) and (beta _t) are estimated by means of the RM filter10 as$$begin{aligned} left. begin{aligned} tilde{mu }_t^{RM}&=medbigl (x_{t-k}+ktilde{beta }_t,ldots ,x_{t+k}-ktilde{beta }_tbigr ),\ tilde{beta }_t^{RM}&=med_{i=-k,ldots ,k} Bigl (med_{j=-k,ldots k,j ne i} frac{x_{t+i}-x_{t+j}}{i-j}Bigr ). end{aligned}right. end{aligned}$$
(4)
The only parameter required for the application of the RM filter is k, which controls how many neighbouring points are included in the estimation of (mu _t). Its choice depends not only on the time series characteristics, but also on the situations a procedure needs to handle. For despiking purposes, k has to be chosen as a trade-off problem between the duration of periods in which trends can be assumed to be approximately linear and the maximum number of consecutive outliers the estimator allows to resist before returning biased results.Results of previous studies18 for the evaluation of the RM filter performance in the removal of patches of impulsive noise showed that the RM resists up to 30% subsequent outliers without being substantially affected. Therefore, the minimal window width should be larger than at least three times the maximal length of outlier patches to be removed.To this end, the optimal time window width selection is carried out through a preliminary analysis of the data distribution. Specifically, the time series is subject to a preliminary de-trending procedure, where trend is approximated by a 5-degree polynomial function whose parameters are estimated via iterated re-weighted least squares (IWLS) regression. The optimal window width is then set equal to 4 times the maximum number of values exceeding (pm 3cdot s_g) in 30 s intervals, where (s_g) is the (global) standard deviation estimated by the (Q_n) estimator on de-trended data. To prevent cases where few or no data exceed the threshold values, a minimum window width of 5 s is imposed (i.e. 51 time steps for data sampled at 10 Hz acquisition frequency).
({{Q}}_n) scale estimatorBeyond the ability of the filter adopted for signal extraction, the effectiveness of a despiking strategy depends also on the robustness of the scale parameter, (sigma _t), which is of fundamental importance for the outlier scores derivation. Raw EC time series cannot be assumed to be identically distributed as variability may vary over time as the effect of changes in turbulence regimes and heterogeneity of the flux footprint area. In such situations, global estimates of the scale parameter are unrepresentative of the local variability. Consequently, the spike detection procedure becomes ineffective. To cope with this feature, the scale parameter (sigma _t) was estimated in rolling time windows whose width was set equal to those adopted for the signal extraction. As a robust estimates of (sigma _t), we used the (Q_n) estimator12$$begin{aligned} Q_n=2.2219{|x_i-x_j|;i0), then the process (X_t) is said to be integrated of order d, meaning that (X_t) needs to be differenced d times to achieve stationarity. To allow heteroskedasticity, we assume that (varepsilon _t= sigma _t e_t), where (e_t) is a sequence of independently and identically distributed variables with mean 0 and variance 1 and (sigma _t^2) is the conditional variance allowed to vary with time.The latter was simulated by means of a CGARCH process, which can be written as:$$begin{aligned} left. begin{aligned} sigma _t^2&=q_t + sum _{i=1}^r alpha _i (varepsilon _{t-i}^2 – q_{t-i}) + sum _{j=1}^s beta _j (sigma _{t-j}^2 -q_{t-j})\ q_t&=omega + eta _{11} q_{t-1} + eta _{21} (varepsilon _{t-1}^2 – sigma _{t-1}^2), end{aligned}right. end{aligned}$$
(7)
where (omega), (alpha _i), (beta _j), (eta _{11}), (eta _{21}) are strictly positive coefficients; (q_t) is the permanent (long-run) component of the conditional variance allowed to vary with time following first order autoregressive type dynamics. The difference between the conditional variance and its trend, (sigma _{t}^2 – q_{t}), is the transitory (short-run) component of the conditional variance. The conditions for the non-negativity estimation of the conditional variance23 are related to the stationary conditions that (alpha _i + beta _j < 1) and that (eta _{11} < 1) (such quantities provide a measure of the persistence of the transitory and permanent components, respectively).Model order specification and parameter estimation were performed by analyzing real EC data (more detail are provided in the “Results and discussion” section). With this modelling framework, we simulated 18,000 values as in EC raw data sampled at 10 Hz scanning frequency within a 30-min interval. Simulations were executed in the R v.4.0.2 programming environment by using the tools implemented in the rugarch package24.Once simulated, synthetic time series were intentionally corrupted with 180 spiky data points (1% for a sample size of 18000). Two macro-scenarios were considered. In the first scenario (S1), isolated or consecutive spike events of short duration were generated. In particular, 180 spike locations were randomly selected in such a way to obtain 30 single spikes, 30 spikes as double (consecutive) events, and 30 spikes as triple (consecutive) events. In the second scenario (S2), instead, time series were contaminated by impulsive peaks of longer duration. To this end, spike locations were carried out by randomly selecting five blocks of 50 consecutive data points. Once located, spikes were generated by multiplying the corresponding time series values (after mean removal) for a factor 10 in such a way to have magnitude similar to those commonly encountered on real, observed EC data. To simulate consecutive spike events as imposed by S2 scenario, generated spiky data points were taken in absolute term. Each scenario was permuted 99 times.MetricsThe ability of the despiking algorithms was assessed by comparing the number of artificial spikes inserted into the time series with the number of spikes identified by the method. More particularly, by referring to the (2times 2) confusion matrix as reported in Table 1, a valid despiking procedure maximizes decisions of type true positive (TP) while, at the same time, keeping decisions of the types false negative (FN) and false positive (FP) at the lowest levels possible. This trade-off can be measured in terms of Precision and Recall, which are commonly used for measuring the effectiveness of set-based retrieval25. For any given threshold value, the Precision is defined as the fraction of reported spikes that truly turn out to be spikes:$$begin{aligned} text {Precision}=frac{text {TP}}{text {TP}+text {FP}}, end{aligned}$$ (8) while the Recall is correspondingly defined as the fraction of ground-truth spikes that have been reported as spikes:$$begin{aligned} text {Recall}=frac{text {TP}}{text {TP}+text {FN}}. end{aligned}$$ (9) Table 1 Confusion matrix.Full size tableAs a measure that combines Precision and Recall, we consider the balanced F1-Score, which is the harmonic mean of the two indices above-mentioned, and given by:$$begin{aligned} text {F1-Score}=2 cdot frac{text {Precision} cdot text {Recall}}{text {Precision} + text {Recall}}. end{aligned}$$ (10) We have (0le text {F1-Score} le 1) where 0 implies that no spikes are detected and 1 indicates that all, and only, the spikes are detected. The closer to 1 the F1-Score index, the greater the effectiveness of the despiking method.In addition to the previous outlined metrics, a comparison between variances of (simulated) uncorrupted time series and the one estimated after the application of the despiking procedure has been performed.For an overall evaluation of the performance of the despiking algorithms, the Friedman test26 using a significance level (alpha =0.05), followed by a post-hoc test based on the procedure introduced in Nemenyi27 was applied. The Friedman test is a non-parametric statistical test, equivalent to repeated-measures ANOVA, which can be used to compare the performances of several algorithms28. The null hypothesis of the Friedman test is that there are no significant differences between performances of all the considered algorithms. Provided that significant differences were detected by the Friedman test (that is the null hypothesis is rejected) the Nemenyi test can be used for pairwise multiple comparisons of the considered algorithms. Nemenyi test is similar to the post-hoc Tukey test for ANOVA, and its output consists of a critical difference (CD) threshold. In order to do that, ranks are assigned to algorithms. For each data set, the algorithm with the best performance gets the lowest (best) average rank. The mean performance of two despiking algorithms is judged to be signifycantly different if the corresponding average ranks differ by at least the critical difference (the graphical output of Nemenyi test was implemented using tools provided in the R package tsutils (https://CRAN.R-project.org/package=tsutils)). More

The development of massive sequencing has provided a relatively inexpensive method to obtain the transcriptome of a species. Taking advantage of this technique, we used a previously obtained transcriptome of P. acuta to identify 18 genes related to different pathways of interest in ecotoxicology and then examined how exposure to phthalates changed the transcription of these genes. The processes of interest include DNA repair, the stress response, detoxification, apoptosis, immunity, energy reserves, and lipid transportation. There is a growing interest in combining ecologically relevant endpoints with biochemical and molecular parameters to seek a more integrative analysis. In this sense, increasing the number of described genes will allow for the design of standard arrays that could be used in combination with toxicity tests. In this way, initiatives such as the Adverse Outcome Pathway wiki24 will increase its relevance in assessing old and new compounds and provide putative mechanisms of action to explain the differences to the animals’ specific physiology. Furthermore, increasing knowledge at the molecular level in P. acuta supports its use as a representative of freshwater gastropods in toxicity analysis. There is a lack of model freshwater mollusks, which is one of the animal groups whose pollution response is currently less known.The 18 newly identified genes evaluated in this work show homology with those previously described in other species, as expected, mainly with the freshwater snail Biomphalaria glabrata, which belongs to the Planorbidae family. rad21 and rad50 are both involved in DNA repair: rad21 is an essential gene encoding a DNA double-strand break repair protein21, and rad50 is a member of the protein complex MRN (including Mre11, RAD50, and Nbs1) that functions in DNA double-strand break repair to recognize and process DNA ends as well as a signal for cell cycle arrest25. There is very little information about these genes in mollusks, with only one report in Crassostrea gigas for rad5026. The relevance of these genes is that their detection can be combined with other methodologies, such as the comet assay, to perform an integrated study to determine whether a compound is genotoxic and whether the organism has the ability to compensate for the damage.The Cat and SOD Mn genes allow us to evaluate the status of oxidative stress. Oxidative stress analysis is usually focused on biochemical parameters, such as enzyme activity. However, it should also include a transcriptional activity study because it can provide additional information about the mid- and long-term responses. Protein turnover can also be relevant in the response, especially in chronic exposure to toxicants. Detoxification mechanisms are also important to assess the response to toxicants. GST activity is one of the most used methods to assess detoxification27, but it does not differentiate between the members involved. The situation is similar regarding cytochrome P450s, which show high diversity with many roles in the cell28. Our identification of the Cyp72a15 gene increases the number of cytochromes 450 s described in P. acuta. Evaluating changes in these genes can help to elucidate how the organism can process the toxicants.The sHSP17.9 and HSC70-4 genes extend the battery of genes available to assess the stress response of P. acuta. sHSP17.9 is difficult to match with other species’ genes because while they all have an alpha-crystallin domain, there is no other sequence that presently allows for homology to be established. Additional functional studies will help to search for homology. It is worth mentioning that HIF1α offers a new aspect of stress related to hypoxia29. The stress response mainly focuses on the canonical heat shock proteins, so other mechanisms involved in specific stresses, such as hypoxia, are usually neglected. With the identification of HIF1a in P. acuta, researchers can evaluate the effect of a toxicant on oxygen intake in this species.The remaining identified genes allow for the analysis of pathways that can also be altered by toxicants, like apoptosis (AIF3), the immune system (ApA), energy reserves (PYGL), and lipid transport (ORP8). To our knowledge, in this study these genes have been analyzed for the first time concerning pollution in freshwater mollusks. The last three genes, DNMT1, KATB6, and HDAC1, are involved in epigenetic mechanisms. There is increasing evidence that epigenetic regulation is one of the long-term effects of toxicants. However, the genes involved in this process in invertebrates are still poorly represented in toxicity analysis. The description of these three genes opens the possibility of analyzing their role in the epigenetic response and its relevance in the transgenerational effects that have started to be described with different toxicants30,31,32.Plastics in the environment are a growing problem. During the degradation process, the polymers themselves and the compounds used as additives, including phthalates, are released. Hence, the presence of phthalates is increasing in the environment5,33,34. We analyzed three phthalates in this work, namely BBP, DEP, and DEHP; they showed a differential impact in P. acuta. DEP and DEHP, did not alter any of the mRNA levels. Researchers have described previously that both phthalates can alter the physiology of invertebrates16,35,36,37,38, including mollusks39,40,41. Other phthalates can also alter development and growth, which could be related to the endocrine-disrupting activity described for those chemicals. The molecular mechanisms involved are still under investigation, but some data are available. In the clam Venerupis philippinarum, DEHP alters the immune response40. In H. diversicolor, DBP affects oxidative stress, lipid and energy metabolism, and osmoregulation17. In other invertebrates, including Chironomus riparius42, Drosophila melanogaster43, and Caenorhabditis elegans15, phthalates alter endocrine pathways. The changes affect the ecdysone response as well as the expression of insulin-like peptide. Other pathways are also affected by phthalates, such as oxidative stress and detoxification routes44 and the stress response14. Finally, in C. elegans, exposure to environmentally relevant concentrations of diethylhexyl phthalate produces genomic instability by altering the expression of genes involved in DNA repair during meiosis37. It is clear then that phthalates can have a broad spectrum of actions in the cell, with a significant alteration of metabolism but primarily affecting oxidative stress and the endocrine system.The previous studies performed in mollusks have revealed alterations in several physiological processes; the analyzed molecular mechanisms mainly involved oxidative stress and immunity17,41. A recent review of the impact of phthalates on aquatic animals summarizes the effects observed, suggesting that activation of the detoxification system (cytochrome P450s) and endocrine system receptors of aquatic animals cause oxidative stress, metabolic disorders, endocrine disorders, and immunosuppression8. It would activate a cascade response that could cause genotoxicity and cell apoptosis, resulting in the disruption of growth and development. Considering this, the absence of a response observed in P. acuta exposed to DEP and DEHP is striking. The differences observed can be assigned to the type of analysis (molecular vs. physiological), the exposure time (1 week vs. a few hours or days), the concentration used (μg/L vs. mg/L), and evidently, the species used. Additional research will help elucidate the differential response in P. acuta compared with other organisms. However, it is essential to highlight that the obtained results suggest that P. acuta can manage the environmentally relevant doses of DEP and DEHP used in this work. This species may be less sensitive to these phthalates, but this eventually will require further research, including the use of other methodological approaches, to confirm it.In contrast to DEP and DEHP, BBP showed a marked effect: it increased the mRNA levels of almost all the analyzed genes. It is essential to consider that most studies on invertebrates that involve transcriptional activity analysis use arthropods and short exposure times14,44,45,46. Limited data are available on mollusks and, usually, they are marine representatives40,47. To our knowledge, this is the first study on a freshwater snail that shows that BBP can produce a substantial effect on cell metabolism. Several of the altered pathways can explain, in some way, the effects observed in other organisms, like DNA repair by the alteration of rad21 and rad50, which are related to DNA damage, or the alteration of the genes involved in histone and DNA modification (KAT6B, HDAC1, and DNMT1), which are related to epigenetic regulation. Apoptosis, which phthalates can also alter, also seems to be modulated in P. acuta by altering the AIF3 and the casp3 genes. Furthermore, the three phases of the detoxification could be acting since the genes tested (three cytochrome P450s, three GSTs, and MRP-1) were upregulated.Genes involved in oxidative stress and the stress response were also altered, as shown by the changes in the mRNA levels of Cat, SODs, stress proteins, and the hypoxia-related transcription factor genes. These changes support the alteration of oxidative stress, the stress response, and detoxification, backing previous analysis and adding new insight about the mechanisms involved in modulating these processes. In this sense, the absence of changes in GSTm1 supports a differential role for each GST family member in the response to toxicants. The altered acetylcholinesterase mRNA level also suggests effects in the nervous system, requiring additional research to elucidate the damage to the central nervous system. Finally, the alteration of PYGL, ApA, and ORP8, involved in energy metabolism, immunity, and lipid transport, respectively, shows that P. acuta responds to BBP in a way that has been observed in other organisms. In summary, the present gene profile obtained in response to BBP in P. acuta supports the proposed mechanisms and cellular processes in studies with other animals8. Immunity, oxidative stress, the stress response, detoxification, apoptosis, epigenetic modulation, DNA repair, lipid metabolism, and energy metabolism are modulated. The nervous system could also be affected. Of note, some genes showed differences in transcription based on the phthalate concentration. These findings suggest there are subtle differences, and additional kinetic analysis is required to elucidate early and late activated genes and the relevance of the damage for the population’s future.The obtained results are in line with previous studies in other organisms, which have confirmed that BBP can induce different types of damage such as apoptosis48, genotoxicity49, oxidative stress50, stress response activation45, or endocrine disruption14. Although there are studies in invertebrates showing the impact on development and other physiological processes39,51, most of them did not focus on the putative mode of action, with only a few of them trying to delve into the response mechanisms. Here we have shown that BBP can extensively affect the cell transcriptional activity in P. acuta. These results could be considered to reflect specific alterations on these pathways. This scenario would mean that BBP is the most active phthalate in P. acuta, with a broad spectrum of action and a potential effect on many pathways. However, the more probable picture is something that has been recently proposed: alterations in the oxidative stress response and the endocrine system cause a cascade of responses that affect different pathways and ultimately block growth and development8. It is relevant to keep in mind that BBP is a known endocrine disruptor47. A recent study in Daphnia magna provides some insight. Specifically, RNA-Seq revealed that genes involved in signal transduction, cell communication, and embryonic development were significantly down-regulated, while those related to biosynthesis, metabolism, cell homeostasis, and redox homeostasis were remarkably upregulated upon BBP exposure46. Although the organism and the stage analyzed are different from our study, those results support the idea that BBP can simultaneously alter multiple pathways, and it fits better with the regulatory role of the endocrine system and the extensive affection by oxidative stress.As stated before, the results obtained in this work show that DEP and DEHP had no apparent effect to P. acuta after 1 week exposure to environmentally relevant concentrations. However, BBP showed a strong effect. The difference in response could be due to several reasons that need to be explored in future work. One possibility is the structure of each compound. In this sense, BBP has two benzene rings while DEP and DEHP have only one. This factor could determine the biological activity of these compounds. Another possibility is that DEP and DEHP have effects earlier than the time studied, and the cell returned to the basal state, being able to process and remove the compounds. Finally, it cannot be dismissed that DEP and DEHP are not toxic to P. acuta, at least at environmentally relevant concentrations. In any case, BBP alters the metabolism of this species and produces a broad impact on different pathways. Additional research should be done in P. acuta and other freshwater species to determine the impact on organisms based on the freshwater ecosystem food web. More