Philippa Kaur

Poloczanska, E. S. et al. Global imprint of climate change on marine life. Nat. Clim. Change 3, 919–925 (2013).Article 
ADS 

Google Scholar 
Wiens, J. J. Climate-related local extinctions are already widespread among plant and animal species. PLoS Biol. 14, e2001104 (2016).Article 

Google Scholar 
Bellard, C., Bertelsmeier, C., Leadley, P., Thuiller, W. & Courchamp, F. Impacts of climate change on the future of biodiversity. Ecol. Lett. 15, 365–377 (2012).Article 

Google Scholar 
Assis, J., Serrão, E. A., Duarte, C. M., Fragkopoulou, E. & Krause-Jensen, D. Major expansion of marine forests in a warmer Arctic. Front. Mar. Sci. 9, 850368 (2022).Article 

Google Scholar 
Assis, J. et al. Major shifts at the range edge of marine forests: The combined effects of climate changes and limited dispersal. Sci. Rep. 7(44348), 1–10 (2017).CAS 

Google Scholar 
O’Leary, J. K. et al. The resilience of marine ecosystems to climatic disturbances. BioScience. https://doi.org/10.1093/biosci/biw161 (2017).Article 

Google Scholar 
Steneck, R. S. et al. Kelp forest ecosystems: Biodiversity, stability, resilience and future. Environ. Conserv. 29, 436–459 (2002).Article 

Google Scholar 
Filbee-Dexter, K. & Scheibling, R. E. Detrital kelp subsidy supports high reproductive condition of deep-living sea urchins in a sedimentary basin. Aquat. Biol. 23, 71–86 (2014).Article 

Google Scholar 
Filbee-Dexter, K. Ocean forests hold unique solutions to our current environmental crisis. One Earth https://doi.org/10.1016/j.oneear.2020.05.004 (2020).Article 

Google Scholar 
Krumhansl, K. A. & Scheibling, R. E. Production and fate of kelp detritus. Mar. Ecol. Prog. Ser. https://doi.org/10.3354/meps09940 (2012).Article 

Google Scholar 
Edwards, M. S. & Hernández-Carmona, G. Delayed recovery of giant kelp near its southern range limit in the North Pacific following El Niño. Mar. Biol. 147, 273–279 (2005).Article 

Google Scholar 
Cavanaugh, K. C., Reed, D. C., Bell, T. W., Castorani, M. C. N. & Beas-Luna, R. Spatial variability in the resistance and resilience of giant kelp in southern and Baja California to a multiyear heatwave. Front. Mar. Sci. https://doi.org/10.3389/fmars.2019.00413 (2019).Article 

Google Scholar 
Butler, C. L., Lucieer, V. L., Wotherspoon, S. J. & Johnson, C. R. Multi-decadal decline in cover of giant kelp Macrocystis pyrifera at the southern limit of its Australian range. Mar. Ecol. Prog. Ser. 653, 1–18 (2020).Article 
ADS 

Google Scholar 
Martínez, B. et al. Distribution models predict large contractions of habitat-forming seaweeds in response to ocean warming. Divers. Distrib. 24, 1350–1366 (2018).Article 

Google Scholar 
Bell, T. W., Allen, J. G., Cavanaugh, K. C. & Siegel, D. A. Three decades of variability in California’s giant kelp forests from the Landsat satellites. Remote Sens. Environ. 238, 110811 (2020).Article 
ADS 

Google Scholar 
Mann, M. E. & Emanuel, K. A. Atlantic Hurricane trends linked to climate change. Eos 87, 233–241 (2006).Article 
ADS 

Google Scholar 
Jensen, J. R., Estes, J. E. & Tinney, L. Remote sensing techniques for kelp surveys. Photogramm. Eng Remote Sens. 46, 743–755 (1980).
Google Scholar 
Cavanaugh, K. C. et al. A review of the opportunities and challenges for using remote sensing for management of surface-canopy forming kelps. Front. Mar. Sci. https://doi.org/10.3389/fmars.2021.753531 (2021).Article 

Google Scholar 
Cavanaugh, K. C., Siegel, D. A., Reed, D. C. & Dennison, P. E. Environmental controls of giant-kelp biomass in the Santa Barbara Channel, California. Mar. Ecol. Prog. Ser. 429, 1–17 (2011).Article 
ADS 

Google Scholar 
Kadhim, M. A. & Abed, M. H. Convolutional neural network for satellite image classification. Stud. Comput. Intell. 830, 165–178 (2020).Article 

Google Scholar 
Segal-Rozenhaimer, M., Li, A., Das, K. & Chirayath, V. Cloud detection algorithm for multi-modal satellite imagery using convolutional neural-networks (CNN). Remote Sens. Environ. 237, 111446 (2020).Article 
ADS 

Google Scholar 
Canonico, G. et al. Global observational needs and resources for marine biodiversity. Front. Mar. Sci. 6, 367 (2019).Article 

Google Scholar 
LeCun, Y., Bengio, Y. & Hinton, G. Deep learning. Nature 521, 436–444 (2015).Article 
ADS 
CAS 

Google Scholar 
Yu, L. & Gong, P. Google Earth as a virtual globe tool for Earth science applications at the global scale: Progress and perspectives. Int. J. Remote Sens. 33, 3966–3986 (2012).Article 

Google Scholar 
Guirado, E., Tabik, S., Rivas, M. L., Alcaraz-Segura, D. & Herrera, F. Whale counting in satellite and aerial images with deep learning. Sci. Rep. 9, 14259 (2019).Article 
ADS 

Google Scholar 
Borowicz, A. et al. Aerial-trained deep learning networks for surveying cetaceans from satellite imagery. PLoS ONE 14, 1–15 (2019).Article 

Google Scholar 
Lorencin, I., Anđelić, N., Mrzljak, V. & Car, Z. Marine objects recognition using convolutional neural networks. Nase More 66, 112–119 (2019).Article 

Google Scholar 
Ridge, J. T., Gray, P. C., Windle, A. E. & Johnston, D. W. Deep learning for coastal resource conservation: Automating detection of shellfish reefs. Remote Sens. Ecol. Conserv. 6, 431–440 (2020).Article 

Google Scholar 
Wang, Y. et al. Machine learning-based ship detection and tracking using satellite images for maritime surveillance. J. Ambient Intell. Smart Environ. 13, 361–371 (2021).Article 

Google Scholar 
Han, Q., Yin, Q., Zheng, X. & Chen, Z. Remote sensing image building detection method based on Mask R-CNN. Complex Intell. Syst. https://doi.org/10.1007/s40747-021-00322-z (2021).Article 

Google Scholar 
Girshick, R. Fast R-CNN. In 2015 IEEE International Conference on Computer Vision (ICCV) 1440–1448. https://doi.org/10.1109/ICCV.2015.169 (2015).Ren, S., He, K., Girshick, R. & Sun, J. Faster R-CNN: Towards real-time object detection with region proposal networks. IEEE Trans Pattern Anal. Mach. Intell. 39, 28 (2017).Article 

Google Scholar 
Shelhamer, E., Long, J. & Darrell, T. Fully convolutional networks for semantic segmentation. IEEE Trans. Pattern Anal. Mach. Intell. 39, 3431–3440 (2017).Article 

Google Scholar 
He, K., Gkioxari, G., Dollár, P. & Girshick, R. Mask R-CNN. In Proceedings of the IEEE international Conference on Computer Vision (2017).Arafeh-Dalmau, N. et al. Extreme Marine Heatwaves alter kelp forest community near its equatorward distribution limit. Front. Mar. Sci. 6, 1–18 (2019).Article 
ADS 

Google Scholar 
Nie, X., Duan, M., Ding, H., Hu, B. & Wong, E. K. Attention Mask R-CNN for ship detection and segmentation from remote sensing images. IEEE Access 8, 9325–9334 (2020).Article 

Google Scholar 
Abdulla, W. Mask R-CNN for object detection and instance segmentation on Keras and TensorFlow. GitHub Repository (2017).Fragkopoulou, E. et al. Global biodiversity patterns of marine forests of brown macroalgae. Glob. Ecol. Biogeogr. https://doi.org/10.1111/geb.13450 (2022).Article 

Google Scholar 
Markham, B. L., Storey, J. C., Williams, D. L. & Irons, J. R. Landsat sensor performance: History and current status. IEEE Trans. Geosci. Remote Sens. https://doi.org/10.1109/TGRS.2004.840720 (2004).Article 

Google Scholar 
Gorelick, N. et al. Google Earth Engine: Planetary-scale geospatial analysis for everyone. Remote Sens. Environ. 202, 18–27 (2017).Article 
ADS 

Google Scholar 
Aghamohamadnia, M. & Abedini, A. A morphology-stitching method to improve Landsat SLC-off images with stripes. Geodesy Geodyn. 5, 27–33 (2014).Article 

Google Scholar 
Houskeeper, H. F. et al. Automated satellite remote sensing of giant kelp at the Falkland Islands (Islas Malvinas). PLoS ONE 17, e0257933 (2022).Article 
CAS 

Google Scholar 
Mantha, K. B. et al. From Fat Droplets to Floating Forests: Cross-Domain Transfer Learning Using a PatchGAN-Based Segmentation Model (2022).Finger, D. J. I., McPherson, M. L., Houskeeper, H. F. & Kudela, R. M. Mapping bull kelp canopy in northern California using Landsat to enable long-term monitoring. Remote Sens. Environ. 254, 112243 (2021).Article 
ADS 

Google Scholar 
Siegel, D. A., Wang, M., Maritorena, S. & Robinson, W. Atmospheric correction of satellite ocean color imagery: The black pixel assumption. Appl. Opt. 39, 3582–3591 (2000).Article 
ADS 
CAS 

Google Scholar 
Loisel, H., Nicolas, J. M., Sciandra, A., Stramski, D. & Poteau, A. Spectral dependency of optical backscattering by marine particles from satellite remote sensing of the global ocean. J. Geophys. Res. Oceans https://doi.org/10.1029/2005JC003367 (2006).Article 

Google Scholar 
Hansen, M. C. et al. High-resolution global maps of 21st-century forest cover change. Science 342, 850–853 (2013).Article 
ADS 
CAS 

Google Scholar 
Dutta, A. & Zisserman, A. The VIA annotation software for images, audio and video. In MM 2019: Proceedings of the 27th ACM International Conference on Multimedia. https://doi.org/10.1145/3343031.3350535 (2019).Pfister, C. A., Berry, H. D. & Mumford, T. The dynamics of Kelp Forests in the Northeast Pacific Ocean and the relationship with environmental drivers. J. Ecol. 106, 1520–1533 (2018).Article 

Google Scholar 
Cavanaugh, K. C., Cavanaugh, K. C., Bell, T. W. & Hockridge, E. G. An automated method for mapping giant kelp canopy dynamics from UAV. Front. Environ. Sci. 8, 587354 (2021).Article 

Google Scholar 
Castorani, M. C. N. et al. Connectivity structures local population dynamics: A long-term empirical test in a large metapopulation system. Ecology 96, 3141–3152 (2015).Article 

Google Scholar 
Irmak, E. Implementation of convolutional neural network approach for COVID-19 disease detection. Physiol. Genom. 52, 590–601 (2020).Article 
CAS 

Google Scholar 
Assis, J., Araújo, M. B. & Serrão, E. A. Projected climate changes threaten ancient refugia of kelp forests in the North Atlantic. Glob. Change Biol. 24, 1365–2486 (2017).
Google Scholar 
Cao, C. et al. An improved faster R-CNN for small object detection. IEEE Access 7, 106838–106846 (2019).Article 

Google Scholar 
Konar, J., Khandelwal, P. & Tripathi, R. Comparison of various learning rate scheduling techniques on convolutional neural network. In 2020 IEEE International Students’ Conference on Electrical, Electronics and Computer Science, SCEECS 2020. https://doi.org/10.1109/SCEECS48394.2020.94 (2020).LeCun, Y., Bottou, L., Bengio, Y. & Haffner, P. Gradient-based learning applied to document recognition. Proc. IEEE 86, 2278–2324 (1998).Article 

Google Scholar 
Johnson, J. W. Automatic nucleus segmentation with mask-RCNN. Adv. Intell. Syst. Comput. 944, 399–407 (2020).
Google Scholar 
Lin, T. Y. et al. Microsoft COCO: Common objects in context. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Vol. 8693 LNCS (2014).McKnight, P. E. & Najab, J. Mann-Whitney U Test. Corsini Encycl. Psychol. https://doi.org/10.1002/9780470479216.corpsy0524 (2010).Article 

Google Scholar 
R Development Core Team. R: A Language and Environment for Statistical Computing (R Foundation for Statistical Computing, 2021).
Google Scholar 
Haklay, M. & Weber, P. OpenStreet map: User-generated street maps. IEEE Pervasive Comput. 7, 12–18 (2008).Article 

Google Scholar 
Wäldchen, J. & Mäder, P. Machine learning for image based species identification. Methods Ecol. Evol. https://doi.org/10.1111/2041-210X.13075 (2018).Article 
MATH 

Google Scholar 
Weinstein, B. G. A computer vision for animal ecology. J. Anim. Ecol. https://doi.org/10.1111/1365-2656.12780 (2018).Article 

Google Scholar 
Chilson, C. et al. Automated detection of bird roosts using NEXRAD radar data and Convolutional Neural Networks. Remote Sens. Ecol. Conserv. 5, 20–32 (2019).Article 

Google Scholar 
O’Gara, S. & McGuinness, K. Comparing data augmentation strategies for deep image classification. Ir. Mach. Vis. Image Process. Conf. https://doi.org/10.21427/148b-ar75 (2019).Article 

Google Scholar 
Li, W., Chen, C., Zhang, M., Li, H. & Du, Q. Data augmentation for hyperspectral image classification with deep CNN. IEEE Geosci. Remote Sens. Lett. 16, 593–597 (2019).Article 
ADS 

Google Scholar 
Bharati, P. & Pramanik, A. Deep learning techniques—R-CNN to Mask R-CNN: A survey. In Computational Intelligence in Pattern Recognition (eds Das, A. K. et al.) 657–668 (Springer, 2020).Chapter 

Google Scholar 
Li, A. S., Chirayath, V., Segal-Rozenhaimer, M., Torres-Perez, J. L. & van den Bergh, J. NASA NeMO-Net’s convolutional neural network: Mapping marine habitats with spectrally heterogeneous remote sensing imagery. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. 13, 5115–5133 (2020).Article 
ADS 

Google Scholar 
Hamilton, S. L., Bell, T. W., Watson, J. R., Grorud-Colvert, K. A. & Menge, B. A. Remote sensing: generation of long-term kelp bed data sets for evaluation of impacts of climatic variation. Ecology 101, e03031 (2020).Article 

Google Scholar 
Bell, T. W., Cavanaugh, K. C. & Siegel, D. A. Remote monitoring of giant kelp biomass and physiological condition: An evaluation of the potential for the Hyperspectral Infrared Imager (HyspIRI) mission. Remote Sens. Environ. 167, 218–228 (2015).Article 
ADS 

Google Scholar 
Schroeder, S. B., Dupont, C., Boyer, L., Juanes, F. & Costa, M. Passive remote sensing technology for mapping bull kelp (Nereocystis luetkeana): A review of techniques and regional case study. Glob. Ecol. Conserv. https://doi.org/10.1016/j.gecco.2019.e00683 (2019).Article 

Google Scholar 
Kristollari, V. & Karathanassi, V. Convolutional neural networks for detecting challenging cases in cloud masking using Sentinel-2 imagery. Remote Sens. Geoinf. Environ. https://doi.org/10.1117/12.2571111 (2020).Article 

Google Scholar 
Wilson, M. J. & Oreopoulos, L. Enhancing a simple MODIS cloud mask algorithm for the landsat data continuity mission. IEEE Trans. Geosci. Remote Sens. 51, 723–731 (2013).Article 
ADS 

Google Scholar 
Zhuge, X. Y., Zou, X. & Wang, Y. A fast cloud detection algorithm applicable to monitoring and nowcasting of daytime cloud systems. IEEE Trans. Geosci. Remote Sens. 55, 6111–6119 (2017).Article 
ADS 

Google Scholar 
Lin, T. Y. et al. Feature pyramid networks for object detection. In Proceedings: 30th IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2017 (2017).Jacox, M. G. et al. Impacts of the 2015–2016 El Niño on the California Current System: Early assessment and comparison to past events. Geophys. Res. Lett. https://doi.org/10.1002/2016GL069716 (2016).Article 

Google Scholar 
Chavez, F. P. et al. Biological and chemical consequences of the 1997–1998 El Niño in central California waters. Prog. Oceanogr. https://doi.org/10.1016/S0079-6611(02)00050-2 (2002).Article 

Google Scholar 
Tegner, M. J. & El Dayton, P. K. Niño effects on Southern California kelp forest communities. Adv. Ecol. Res. 17, 243–279 (1987).Article 

Google Scholar 
Bartsch, I. et al. Changes in kelp forest biomass and depth distribution in Kongsfjorden, Svalbard, between 1996–1998 and 2012–2014 reflect Arctic warming. Polar Biol. 39, 2021–2036 (2016).Article 

Google Scholar 
Simonson, E. J., Scheibling, R. E. & Metaxas, A. Kelp in hot water: I. Warming seawater temperature induces weakening and loss of kelp tissue. Mar. Ecol. Prog. Ser. https://doi.org/10.3354/meps11438 (2015).Article 

Google Scholar 
Oliver, E. C. J. et al. Projected marine heatwaves in the 21st century and the potential for ecological impact. Front. Mar. Sci. https://doi.org/10.3389/fmars.2019.00734 (2019).Article 

Google Scholar  More

Schilling, A.-M. & Rössner, G. E. The (sleeping) beauty in the beast—a review on the water deer, Hydropotes inermis. Hystrix Ital. J. Mammal. 28, 121–133 (2017).
Google Scholar 
Geist, V. Deer of the World: Their Evolution, Behaviour and Ecology (Stackpole Books, Pennsylvania, 1998).
Google Scholar 
Cooke, A. Muntjac and Water Deer: Natural History, Environmental Impact and Management (Pelagic Publishing Ltd, UK, 2019).Book 

Google Scholar 
Kim, B. J., Lee, B. K. & Kim, Y. J. Korean water deer (National Institute of Ecology, South Korea, 2016).
Google Scholar 
Belyaev, D. A. & Jo, Y.-S. Northernmost finding and further information on water deer Hydropotes inermis in Primorskiy Krai, Russia. Mammalia 85, 71–73 (2021).Article 

Google Scholar 
Harris, R. B. & Duckworth, J. W. Hydropotes inermis. The IUCN Red List of Threatened Species, e.T10329A22163569 (2015).National Institute of Biological Resources. Harmful wildlife. https://species.nibr.go.kr/home/mainHome.do?cont_link=011&subMenu=011016&contCd=011016001 (2015).Hofmann, R. R. Evolutionary steps of ecophysiological adaptation and diversification of ruminants: a comparative view of their digestive system. Oecologia 78, 443–457 (1989).Article 
ADS 
CAS 

Google Scholar 
Guo, G. & Zhang, E. Diet of the Chinese water deer (Hydropotes inermis) in Zhoushan Archipelago, China. Acta Theriol. Sin. 25, 122–130 (2005).
Google Scholar 
Kim, B. J., Lee, N. S. & Lee, S. D. Feeding diets of the Korean water deer (Hydropotes inermis argyropus) based on a 202 bp rbcL sequence analysis. Conserv. Genet. 12, 851–856 (2011).Article 

Google Scholar 
Park, J.-E., Kim, B.-J., Oh, D.-H., Lee, H. & Lee, S.-D. Feeding habit analysis of the Korean water deer. Korean J. Environ. Ecol. 25, 836–845 (2011).
Google Scholar 
Kim, J., Joo, S. & Park, S. Diet composition of Korean water deer (Hydropotes inermis argyropus) from the Han River Estuary Wetland in Korea using fecal DNA. Mammalia 85, 487–493 (2021).Article 

Google Scholar 
Hofmann, R., Kock, R. A., Ludwig, J. & Axmacher, H. Seasonal changes in rumen papillary development and body condition in free ranging Chinese water deer (Hydropotes inermis). J. Zool. 216, 103–117 (1988).Article 

Google Scholar 
Nielsen, J. M., Clare, E. L., Hayden, B., Brett, M. T. & Kratina, P. Diet tracing in ecology: Method comparison and selection. Methods Ecol. Evol. 9, 278–291 (2018).Article 

Google Scholar 
Birnie-Gauvin, K., Peiman, K. S., Raubenheimer, D. & Cooke, S. J. Nutritional physiology and ecology of wildlife in a changing world. Conserv. Physiol. 5, cox030 (2017).Article 

Google Scholar 
Taberlet, P., Coissac, E., Pompanon, F., Brochmann, C. & Willerslev, E. Towards next-generation biodiversity assessment using DNA metabarcoding. Mol. Ecol. 21, 2045–2050 (2012).Article 
CAS 

Google Scholar 
Glenn, T. C. Field guide to next-generation DNA sequencers. Mol. Ecol. Resour. 11, 759–769 (2011).Article 
CAS 

Google Scholar 
Nichols, R. V., Åkesson, M. & Kjellander, P. Diet assessment based on rumen contents: A comparison between DNA metabarcoding and macroscopy. PLoS ONE 11, e0157977 (2016).Article 

Google Scholar 
Pompanon, F. et al. Who is eating what: diet assessment using next generation sequencing. Mol. Ecol. 21, 1931–1950 (2012).Article 
CAS 

Google Scholar 
Kumari, P. et al. DNA metabarcoding-based diet survey for the Eurasian otter (Lutra lutra): Development of a Eurasian otter-specific blocking oligonucleotide for 12S rRNA gene sequencing for vertebrates. PLoS ONE 14, e0226253 (2019).Article 
CAS 

Google Scholar 
Iwanowicz, D. D. et al. Metabarcoding of fecal samples to determine herbivore diets: A case study of the endangered Pacific pocket mouse. PLoS ONE 11, e0165366 (2016).Article 

Google Scholar 
Berry, T. E. et al. DNA metabarcoding for diet analysis and biodiversity: A case study using the endangered Australian sea lion (Neophoca cinerea). Ecol. Evol. 7, 5435–5453 (2017).Article 

Google Scholar 
Ford, M. J. et al. Estimation of a killer whale (Orcinus orca) population’s diet using sequencing analysis of DNA from feces. PLoS ONE 11, e0144956 (2016).Article 

Google Scholar 
Ando, H. et al. Diet analysis by next-generation sequencing indicates the frequent consumption of introduced plants by the critically endangered red-headed wood pigeon (Columba janthina nitens) in oceanic island habitats. Ecol. Evol. 3, 4057–4069 (2013).Article 

Google Scholar 
Kim, E.-K. Behavioral ecology, habitat evaluation and genetic characteristics of water deer (Hydropotes inermis) in Korea. Ph.D. thesis. Kangwon National University (2011).Park, J.-E., Kim, B.-J. & Lee, S.-D. A study of potential of diet analysis in the Korean water deer (Hydropotes inermis argyropus) using polymerase chain reaction-denaturing gradient gel electrophoresis (PCR-DGGE). Korean J. Environ. Ecol. 24, 318–324 (2010).
Google Scholar 
Hollingsworth, P. M. Refining the DNA barcode for land plants. Proc. Natl. Acad. Sci. USA 108, 19451–19452 (2011).Article 
ADS 
CAS 

Google Scholar 
Li, D.-Z. et al. Comparative analysis of a large dataset indicates that internal transcribed spacer (ITS) should be incorporated into the core barcode for seed plants. Proc. Natl. Acad. Sci. USA 108, 19641–19646 (2011).Article 
ADS 
CAS 

Google Scholar 
Park, E. & Nam, M. Changes in land cover and the cultivation area of ginseng in the Civilian Control Zone -Paju City and Yeoncheon County-. Korean J. Environ. Ecol. 27, 507–515 (2013).
Google Scholar 
Cheng, T. et al. Barcoding the kingdom Plantae: new PCR primers for ITS regions of plants with improved universality and specificity. Mol. Ecol. Resour. 16, 138–149 (2016).Article 
CAS 

Google Scholar 
Edgar, R. C. Search and clustering orders of magnitude faster than BLAST. Bioinformatics 26, 2460–2461 (2010).Article 
CAS 

Google Scholar 
Ankenbrand, M. J., Keller, A., Wolf, M., Schultz, J. & Förster, F. ITS2 database V: Twice as much. Mol. Biol. Evol. 32, 3030–3032 (2015).Article 
CAS 

Google Scholar 
Sickel, W. et al. Increased efficiency in identifying mixed pollen samples by meta-barcoding with a dual-indexing approach. BMC Ecol. 15, 20 (2015).Article 

Google Scholar 
Edgar, R. C. Accuracy of taxonomy prediction for 16S rRNA and fungal ITS sequences. PeerJ 6, e4652 (2018).Article 

Google Scholar 
Oksanen, J. et al. vegan: Community ecology package v 2.5–7 (R Foundation, Vienna, Austria, 2020).
Google Scholar 
Hsieh, T., Ma, K. & Chao, A. iNEXT: an R package for rarefaction and extrapolation of species diversity (Hill numbers). Methods Ecol. Evol. 7, 1451–1456 (2016).Article 

Google Scholar 
Anderson, M. J. A new method for non-parametric multivariate analysis of variance. Austral Ecol. 26, 32–46 (2001).
Google Scholar 
De Cáceres, M. & Legendre, P. Associations between species and groups of sites: indices and statistical inference. Ecology 90, 3566–3574 (2009).Article 

Google Scholar 
Yan, L. ggvenn: Draw venn diagram by ‘ggplot2’ v. 0.1.8 (R Foundation, Vienna, Austria, 2021).Choi, D.-Y. et al. Flora of province Gyonggi-do. Bull. Seoul Nat’l Univ. Arbor. 21, 25–76 (2001).
Google Scholar 
Ko, S. & Shin, Y. Flora of middle part in Gyeonggi Province. Korean J. Plant Res. 22, 49–70 (2009).
Google Scholar 
Lee, S.-K., Ryu, Y. & Lee, E. J. Endozoochorous seed dispersal by Korean water deer (Hydropotes inermis argyropus) in Taehwa Research Forest, South Korea. Glob. Ecol. Conserv. 40, e02325 (2022).Article 

Google Scholar 
Kim, K.-H. & Kang, S.-H. Flora of western civilian control zone (CCZ) in Korea. Korean J. Plant Res. 32, 565–588 (2019).
Google Scholar 
Gyeonggi Tourism Organization. Pyeonghwa-Nuri Trail ecological resource survey. (Paju City, Gyeonggi Province, Korea, 2018).Wickham, H. ggplot2: Elegant Graphics for Data Analysis 2nd edn. (Springer, New York, 2016).Book 
MATH 

Google Scholar 
R Core Team. R: A language and environment for statistical computing (R Foundation, Vienna, Austria, 2020).Pertoldi, C. et al. Comparing DNA metabarcoding with faecal analysis for diet determination of the Eurasian otter (Lutra lutra) in Vejlerne. Denmark. Mammal. Res. 66, 115–122 (2021).Article 

Google Scholar 
Lee, B. Morphological, ecological and DNA taxonomic characteristics of Chinese water deer (Hydropotes inermis Swinhoe). Ph.D. thesis. Chungbuk National University (2003).Wilmshurst, J. F., Fryxell, J. M. & Hudsonb, R. J. Forage quality and patch choice by wapiti (Cervus elaphus). Behav. Ecol. 6, 209–217 (1995).Article 

Google Scholar 
Langvatn, R. & Hanley, T. A. Feeding-patch choice by red deer in relation to foraging efficiency. Oecologia 95, 164–170 (1993).Article 
ADS 

Google Scholar 
Gray, P. B. & Servello, F. A. Energy intake relationships for white-tailed deer on winter browse diets. J. Wildl. Manag. 59, 147–152 (1995).Article 

Google Scholar 
Brown, D. T. & Doucet, G. J. Temporal changes in winter diet selection by white-tailed deer in a northern deer yard. J. Wildl. Manag. 55, 361–376 (1991).Article 

Google Scholar 
Takahashi, H. & Kaji, K. Fallen leaves and unpalatable plants as alternative foods for sika deer under food limitation. Ecol. Res. 16, 257–262 (2001).Article 

Google Scholar 
Bee, J. N. et al. Spatio-temporal feeding selection of red deer in a mountainous landscape. Austral Ecol. 35, 752–764 (2010).Article 

Google Scholar 
Gebert, C. & Verheyden-Tixier, H. Variations of diet composition of red deer (Cervus elaphus L.) in Europe. Mammal. Rev. 31, 189–201 (2001).Article 

Google Scholar 
Cornelis, J., Casaer, J. & Hermy, M. Impact of season, habitat and research techniques on diet composition of roe deer (Capreolus capreolus): a review. J. Zool. 248, 195–207 (1999).Article 

Google Scholar 
Kim, B. J. & Lee, S.-D. Home range study of the Korean water deer (Hydropotes inermis agyropus) using radio and GPS tracking in South Korea: Comparison of daily and seasonal habitat use pattern. J. Ecol. Field Biol. 34, 365–370 (2011).
Google Scholar 
Beier, P. Sex differences in quality of white-tailed deer diets. J. Mammal. 68, 323–329 (1987).Article 

Google Scholar 
Staines, B. W., Crisp, J. M. & Parish, T. Differences in the quality of food eaten by red deer (Cervus elaphus) stags and hinds in winter. J. Appl. Ecol. 19, 65–77 (1982).Article 

Google Scholar 
Koga, T. & Ono, Y. Sexual differences in foraging behavior of sika deer, Cervus nippon. J. Mammal. 75, 129–135 (1994).Article 

Google Scholar 
Han, S.-H., Lee, S.-S., Cho, I.-C., Oh, M.-Y. & Oh, H.-S. Species identification and sex determination of Korean water deer (Hydropotes inermis argyropus) by duplex PCR. J. Appl. Anim. Res. 35, 61–66 (2009).Article 
CAS 

Google Scholar 
You, Z. et al. Seasonal variations in the composition and diversity of gut microbiota in white-lipped deer (Cervus albirostris). PeerJ 10, e13753 (2022).Article 

Google Scholar 
Zhao, W. et al. Metagenomics analysis of the gut microbiome in healthy and bacterial pneumonia forest musk deer. Gene Genom. 43, 43–53 (2021).Article 
CAS 

Google Scholar 
Amato, K. R. et al. Gut microbiome, diet, and conservation of endangered langurs in Sri Lanka. Biotropica 52, 981–990 (2020).Article 

Google Scholar 
Stumpf, R. M. et al. Microbiomes, metagenomics, and primate conservation: New strategies, tools, and applications. Biol. Conserv. 199, 56–66 (2016).Article 

Google Scholar 
Redford, K. H., Segre, J. A., Salafsky, N., del Rio, C. M. & McAloose, D. Conservation and the microbiome. Conserv. Biol. 26, 195–197 (2012).Article 

Google Scholar 
Deagle, B. E. et al. Counting with DNA in metabarcoding studies: How should we convert sequence reads to dietary data?. Mol. Ecol. 28, 391–406 (2019).Article 

Google Scholar 
Corse, E. et al. A from-benchtop-to-desktop workflow for validating HTS data and for taxonomic identification in diet metabarcoding studies. Mol. Ecol. Resour. 17, e146–e159 (2017).Article 
CAS 

Google Scholar 
Alberdi, A. et al. Promises and pitfalls of using high-throughput sequencing for diet analysis. Mol. Ecol. Resour. 19, 327–348 (2019).Article 

Google Scholar 
Nakahara, F. et al. The applicability of DNA barcoding for dietary analysis of sika deer. DNA Barcodes 3, 200–206 (2015).Article 

Google Scholar 
Thomas, A. C., Jarman, S. N., Haman, K. H., Trites, A. W. & Deagle, B. E. Improving accuracy of DNA diet estimates using food tissue control materials and an evaluation of proxies for digestion bias. Mol. Ecol. 23, 3706–3718 (2014).Article 
CAS 

Google Scholar 
Deagle, B. E., Eveson, J. P. & Jarman, S. N. Quantification of damage in DNA recovered from highly degraded samples–a case study on DNA in faeces. Front. Zool. 3, 11 (2006).Article 

Google Scholar 
Coissac, E., Riaz, T. & Puillandre, N. Bioinformatic challenges for DNA metabarcoding of plants and animals. Mol. Ecol. 21, 1834–1847 (2012).Article 
CAS 

Google Scholar 
Estes, J. A. et al. Trophic downgrading of planet Earth. Science 333, 301–306 (2011).Article 
ADS 
CAS 

Google Scholar 
Clare, E. L. Molecular detection of trophic interactions: emerging trends, distinct advantages, significant considerations and conservation applications. Evol. Appl. 7, 1144–1157 (2014).Article 

Google Scholar 
Ramirez, R., Quintanilla, J. & Aranda, J. White-tailed deer food habits in northeastern Mexico. Small Rumin. Res. 25, 141–146 (1997).Article 

Google Scholar 
Anouk Simard, M., Côté, S. D., Weladji, R. B. & Huot, J. Feedback effects of chronic browsing on life-history traits of a large herbivore. J. Anim. Ecol. 77, 678–686 (2008).Article 
CAS 

Google Scholar 
Putman, R. J. & Staines, B. W. Supplementary winter feeding of wild red deer Cervus elaphus in Europe and North America: justifications, feeding practice and effectiveness. Mammal Rev. 34, 285–306 (2004).Article 

Google Scholar 
Milner, J. M., Van Beest, F. M., Schmidt, K. T., Brook, R. K. & Storaas, T. To feed or not to feed? Evidence of the intended and unintended effects of feeding wild ungulates. J. Wildl. Manag. 78, 1322–1334 (2014).Article 

Google Scholar 
Carpio, A. J., Apollonio, M. & Acevedo, P. Wild ungulate overabundance in Europe: contexts, causes, monitoring and management recommendations. Mammal Rev. 51, 95–108 (2021).Article 

Google Scholar 
Cappa, F., Lombardini, M. & Meriggi, A. Influence of seasonality, environmental and anthropic factors on crop damage by wild boar Sus scrofa. Folia Zool. 68, 261–268 (2019).Article 

Google Scholar  More

Stokland, J. N., Siitonen, J. & Jonsson, B. G. Biodiversity in Dead Wood (Cambridge Univ. Press, 2012).Turetsky, M. R. et al. Nat. Geosci. 8, 11–14 (2014).Article 

Google Scholar 
Wenger, S. J., Subalusky, A. L. & Freeman, M. C. Food Webs 18, e00106 (2019).Article 

Google Scholar 
Tomatsuri, M. & Kon, K. Hydrobiologia 790, 225–232 (2017).Article 

Google Scholar 
Henry, L. A. & Roberts, J. M. in Marine Animal Forests (eds Rossi, S. et al.) 235–256 (Springer, 2017).Walton, M. E. M. et al. Sci. Total Environ. 820, 153191 (2022).Article 
CAS 

Google Scholar 
Wolfe, K., Kenyon, T. M. & Mumby, P. J. Coral Reefs 40, 1769–1806 (2021).Article 

Google Scholar 
Kim, H. et al. Glob. Change Biol. 28, 6180–6193 (2022).Jackson, R. B. et al. Annu. Rev. Ecol. Evol. Syst. 48, 419–445 (2017).Article 

Google Scholar 
Pan, Y. et al. Science 333, 988–993 (2011).Article 
CAS 

Google Scholar 
Hedges, J. I., Keil, R. G. & Benner, R. Org. Geochem. 27, 195–212 (1997).Article 
CAS 

Google Scholar 
Lønborg, C. et al. Front. Mar. Sci. 7, 466 (2020).Article 

Google Scholar 
Harden, J. W. et al. Glob. Change Biol. 6, 174–184 (2000).Davidson, E. A. & Janssens, I. A. Nature 440, 165–173 (2006).Article 
CAS 

Google Scholar 
Hugelius, G. et al. Proc. Natl Acad. Sci. USA 117, 20438–20446 (2020).Article 
CAS 

Google Scholar 
Hennige, S. J. et al. Front. Mar. Sci. https://doi.org/10.3389/fmars.2020.00668 (2020).Article 

Google Scholar 
Wolfram, U. et al. Sci. Rep. 12, 8052 (2022).Article 
CAS 

Google Scholar 
Roberts, J. M., Wheeler, A. J. & Freiwald, A. Science 312, 543–547 (2006).Article 
CAS 

Google Scholar 
Mortensen, P. B. & Fosså, J. H. Species diversity and spatial distribution of invertebrates on deep-water Lophelia reefs in Norway. In Proc. 10th Int. Coral Reef Symp. 1849–1868 (ICRS, 2006).Maier, S. R. et al. Deep Sea Res. I 175, 103574 (2021).. More

MAG binning and general featuresFrom the four hot springs, we assembled four associated metagenomes and then binned a total of 42 MAGs. We recovered 12 MAGs from RB10 hot spring, 13 from RB13, 14 from RB32 and 3 from RB108. Out of these 42 MAGs, 7 were of high-quality, 25 of nearly-high quality, 9 of medium quality and 1 of low quality (Table 1) based on metagenomic standards26. The GC% was quite variable, ranging from 25.76 to 70.35% among all MAGs and between 32.15 and 69.21% only among the high- and near high-quality MAGs. With the exception of RB108 from which we only recovered bacterial MAGs, we retrieved both bacterial and archaeal MAGs in the other hot springs. Two thirds of the MAGs (26/42) were assigned to the domain Bacteria and the rest to the domain Archaea (16/42) (Table 2).Table 1 General characteristics of the 42 MAGs obtained from RB10, RB13, RB32 and RB108 samples.Full size tableTable 2 Classification of the MAGs based on the taxonomic classification of GTDB-Tk software (v2.1.0) and the Genome Taxonomy Database (07-RS207 release).Full size tableTaxonomic and phylogenomic analyses of MAGsThe taxonomic affiliation of the MAGs was investigated in detail through the workflow classify of GTDB-Tk (v 2.1.0; GTDB reference tree 07-RS207) (Table 2) and through de novo phylogenomic analyses (Fig. S1a–i). We also tried to classify MAGs on the basis of overall genome relatedness indices (OGRI), which is detailed in supplementary material (Text S1, Table S2, Fig. S2).De novo phylogenomic analyses globally confirmed the positioning of MAGs provided by GTDB-Tk, with high branching support. For Bacteria, GTDB-Tk analyses allowed us to place the MAGs in the following clades: six in the phylum Aquificota from the four different springs, comprising four MAGs belonging to the genus Hydrogenivirga (family Aquificaceae) (RB10-MAG07, RB13-MAG10, RB32-MAG07, RB108-MAG02), and two belonging to the family ‘Hydrogenobaculaceae’ (RB10-MAG12, RB32-MAG11) (Table 2, Fig. S1a). Their closest cultured relatives originated either from hot springs or from deep-sea hydrothermal vents27. Three MAGs from three geothermal springs belonged to the phylum Armatimonadota (RB10-MAG03, RB13-MAG04, RB32-MAG03) and had no close cultured relatives. Seven MAGs have been classified into the phylum Chloroflexota: three MAGs belonging to the genus Thermoflexus from three different springs (RB10-MAG04, RB13-MAG05, RB32-MAG02), one affiliating with the genus Thermomicrobium (RB32-MAG08), one falling into the family Ktedonobacteraceae (RB108-MAG03), one belonging to the class Dehalococcoidia (RB32-MAG04) and another one whose phylogenetic position is more difficult to assert because it is a MAG of medium quality (RB32-MAG14). Six MAGs from four various hot springs belonged to the phylum Deinococcota, and to the genera Thermus (RB10-MAG08, RB10-MAG11, RB13-MAG09, RB32-MAG10, RB108-MAG01) and Meiothermus (RB13-MAG13). One MAG belonged to the family ‘Sulfurifustaceae’ (RB13-MAG01), in the phylum Proteobacteria (Gamma-class). The MAG referenced as RB32-MAG13 was classified into the phylum ‘Patescibacteria’, in the class ‘Paceibacteria’, and was distantly related to MAGs originating from groundwater and from hot springs. Finally, two MAGs from two different springs belonged to the phylum WOR-3, in the Candidatus genus ‘Caldipriscus’ (RB32-MAG12, RB10-MAG09).For Archaea, almost all the MAGs reconstructed in this study, e.g. 15 of the 16 archaeal MAGs, belonged to the phylum Thermoproteota. Among them, four belonged to the genus Ignisphaera (RB10-MAG05, RB13-MAG08, RB13-MAG11, RB32-MAG05), three to the genus Infirmifilum (RB10-MAG06, RB13-MAG03, RB32-MAG09), two to the genus Zestosphaera (RB10-MAG02, RB13-MAG06), three to the family Acidilobaceae (RB10-MAG01, RB13-MAG02, RB32-MAG01) and two to the order Geoarchaeales (RB10-MAG10, RB32-MAG06). Additionally, one belonged to the family Thermocladiaceae (RB13-MAG07). Lastly, the MAG belonging to another phylum (RB13-MAG12) was affiliated with the ‘Aenigmatarchaeota’, class ‘Aenigmatarchaeia’, and was distantly related to MAGs from hot springs and from deep-sea hydrothermal vent sediments28,29.Out of these 42 MAGs, at least 19 MAGs corresponded to different taxa at the taxonomic rank of species or higher according to GTDB (Table 2). Eighteen of them belonged to lineages with several cultivated representatives including the species Thermus thermophilus. 13 new genomic species within the GTDB genera Hydrogenivirga, HRBIN17, Thermoflexus, SpSt-223, CADDYT01, Zestosphaera, Ignisphaera, Infirmifilum, Thermus, Thermus_A, Meiothermus_B, JAHLMO01 and Caldipriscus, and 6 putative new genomic genera belonging to the GTDB families Hydrogenobaculaceae, Acidilobaceae, WAQG01, Thermocladiaceae, Sulfurifustaceae and HR35 could be identified (Table 2). In addition, 9 MAGs belonged to lineages that are predominantly or exclusively known through environmental DNA sequences. Thus, these 42 MAGs comprised a broad phylogenetic range of Bacteria and Archaea at different levels of taxonomic organization, of which a large majority were not reported before.The approaches implemented here were not intended to describe the microbial diversity present in these sources in an exhaustive way or to compare them in a fine way, and cannot allow it because of a 2-year storage at 4 °C. This long storage has probably led to changes in the microbial communities and to the selective loss or enrichment of some taxa. As a result, no analysis of abundance or absence of taxa can be conducted from these metagenomes and the results are discussed taking this bias into account. However, they do provide an overview of the microbial diversity effectively present. If we compare the phylogenetic diversity of the MAGs found in the four hot springs, we can observe that 3 shared phyla (Deinococcota, Aquificota and Chloroflexota: phyla names according to GTDB), 2 shared families (Thermaceae and Aquificaceae), and one shared genus (Hydrogenivirga) were found among the four sources (Fig. 2). In addition, hot springs RB10, RB13 and RB32, that are geographically close ( More

Monte Carlo analysisSeaweed production costs and net costs of climate benefits were estimated on the basis of outputs of the biophysical and technoeconomic models described below. The associated uncertainties and sensitivities were quantified by repeatedly sampling from uniform distributions of plausible values for each cost and economic parameter (n = 5,000 for each nutrient scenario from the biophysical model, for a total of n = 10,000 simulations; see Supplementary Figs. 14 and 15)47,48,49,50,51,52. Parameter importance across Monte Carlo simulations (Fig. 3 and Supplementary Fig. 9) was determined using decision trees in LightGBM, a gradient-boosting machine learning framework.Biophysical modelG-MACMODS is a nutrient-constrained, biophysical macroalgal growth model with inputs of temperature, nitrogen, light, flow, wave conditions and amount of seeded biomass30,53, that we used to estimate annual seaweed yield per area (either in tons of carbon or tons of dry weight biomass per km2 per year)33,34. In the model, seaweed takes up nitrogen from seawater, and that nitrogen is held in a stored pool before being converted to structural biomass via growth54. Seaweed biomass is then lost via mortality, which includes breakage from variable ocean wave intensity. The conversion from stored nitrogen to biomass is based on the minimum internal nitrogen requirements of macroalgae, and the conversion from biomass to units of carbon is based on an average carbon content of macroalgal dry weight (~30%)55. The model accounts for farming intensity (sub-grid-scale crowding) and employs a conditional harvest scheme, where harvest is optimized on the basis of growth rate and standing biomass33.The G-MACMODS model is parameterized for four types of macroalgae: temperate brown, temperate red, tropical brown and tropical red. These types employed biophysical parameters from genera that represent over 99.5% of present-day farmed macroalgae (Eucheuma, Gracilaria, Kappahycus, Sargassum, Porphyra, Saccharina, Laminaria, Macrocystis)39. Environmental inputs were derived from satellite-based and climatological model output mapped to 1/12-degree global resolution, which resolves continental shelf regions. Nutrient distributions were derived from a 1/10-degree resolution biogeochemical simulation led by the National Center for Atmospheric Research (NCAR) and run in the Community Earth System Model (CESM) framework35.Two nutrient scenarios were simulated with G-MACMODS and evaluated using the technoeconomic model analyses described below: the ‘ambient nutrient’ scenario where seaweed growth was computed using surface nutrient concentrations without depletion or competition, and ‘limited nutrient’ simulations where seaweed growth was limited by an estimation of the nutrient supply to surface waters (computed as the flux of deep-water nitrate through a 100 m depth horizon). For each Monte Carlo simulation in the economic analysis, the technoeconomic model randomly selects either the 5th, 25th, 50th, 75th or 95th percentile G-MACMODS seaweed yield map from a normal distribution to use as the yield map for that simulation. Figures and numbers reported in the main text are based on the ambient-nutrient scenario; results based on the limited-nutrient scenario are shown in Supplementary Figures.Technoeconomic modelAn interactive web tool of the technoeconomic model is available at https://carbonplan.org/research/seaweed-farming.We estimated the net cost of seaweed-related climate benefits by first estimating all costs and emissions related to seaweed farming, up to and including the point of harvest at the farm location, then estimating costs and emissions related to the transportation and processing of harvested seaweed, and finally estimating the market value of seaweed products and either carbon sequestered or GHG emissions avoided.Production costs and emissionsSpatially explicit costs of seaweed production ($ tDW−1) and production-related emissions (tCO2 tDW−1) were calculated on the basis of ranges of capital costs ($ km−2), operating costs (including labour, $ km−2), harvest costs ($ km−2) and transport emissions per distance travelled (tCO2 km−1) in the literature (Table 1, Supplementary Tables 1 and 2); annual seaweed biomass (tDW km−2, for the preferred seaweed type in each grid cell), line spacing and number of harvests (species-dependent) from the biophysical model; as well as datasets of distances to the nearest port (km), ocean depth (m) and significant wave height (m).Capital costs were calculated as:$$c_{cap} = c_{capbase} + left( {c_{capbase} times left( {k_d + k_w} right)} right) + c_{sl}$$
(1)
where ccap is the total annualized capital costs per km2, ccapbase is the annualized capital cost per km2 (for example, cost of buoys, anchors, boats, structural rope) before applying depth and wave impacts, kd and kw are the impacts of depth and waviness on capital cost, respectively, each expressed as a multiplier between 0 and 1 modelled using our Monte Carlo method and applied only to grid cells with depth >500 m and/or significant wave height >3 m, respectively, and csl is the total annual cost of seeded line calculated as:$$c_{sl} = c_{slbase} times p_{sline}$$
(2)
where cslbase is the cost per metre of seeded line, and psline is the total length of line per km2, based on the optimal seaweed type grown in each grid cell.Operating and maintenance costs were calculated as:$$c_{op} = c_{ins} + c_{lic} + c_{lab} + c_{opbase}$$
(3)
where cop is the total annualized operating and maintenance costs per km2, cins is the annual insurance cost per km2, clic is the annual cost of a seaweed aquaculture license per km2, clab is the annual cost of labour excluding harvest labour, and copbase is all other operating and maintenance costs.Harvest costs were calculated as:$$c_{harv} = c_{harvbase} times n_{harv}$$
(4)
where charv is the total annual costs associated with harvesting seaweed per km2, charvbase is the cost per harvest per km2 (including harvest labour but excluding harvest transport), and nharv is the total number of harvests per year.Costs associated with transporting equipment to the farming location were calculated as:$$c_{eqtrans} = c_{transbase} times m_{eq} times d_{port}$$
(5)
where ceqtrans is total annualized cost of transporting equipment, ctransbase is the cost to transport 1 ton of material 1 km on a barge, meq is the annualized equipment mass in tons and dport is the ocean distance to the nearest port in km.The total production cost of growing and harvesting seaweed was therefore calculated as:$$c_{prod} = frac{{left( {c_{cap}} right) + left( {c_{op}} right) + left( {c_{harv}} right) + (c_{eqtrans})}}{{s_{dw}}}$$
(6)
where cprod is total annual cost of seaweed production (growth + harvesting), ccap is as calculated in equation (1), cop is as calculated in equation (3), charv is as calculated in equation (4), ceqtrans is as calculated in equation (5) and sdw is the DW of seaweed harvested annually per km2.Emissions associated with transporting equipment to the farming location were calculated as:$$e_{eqtrans} = e_{transbase} times m_{eq} times d_{port}$$
(7)
where eeqtrans is the total annualized CO2 emissions in tons from transporting equipment, etransbase is the CO2 emissions from transporting 1 ton of material 1 km on a barge, meq is the annualized equipment mass in tons and dport is the ocean distance to the nearest port in km.Emissions from maintenance trips to/from the seaweed farm were calculated as:$$e_{mnt} = left( {left( {2 times d_{port}} right) times e_{mntbase} times left( {frac{{n_{mnt}}}{{a_{mnt}}}} right)} right) + (e_{mntbase} times d_{mnt})$$
(8)
where emnt is total annual CO2 emissions from farm maintenance, dport is the ocean distance to the nearest port in km, nmnt is the number of maintenance trips per km2 per year, amnt is the area tended to per trip, dmnt is the distance travelled around each km2 for maintenance and emntbase is the CO2 emissions from travelling 1 km on a typical fishing maintenance vessel (for example, a 14 m Marinnor vessel with 2 × 310 hp engines) at an average speed of 9 knots (16.67 km h−1), resulting in maintenance vessel fuel consumption of 0.88 l km−1 (refs. 28,56).Total emissions from growing and harvesting seaweed were therefore calculated as:$$e_{prod} = frac{{(e_{eqtrans}) + (e_{mnt})}}{{s_{dw}}}$$
(9)
where eprod is total annual emissions from seaweed production (growth + harvesting), eeqtrans is as calculated in equation (7), emnt is as calculated in equation (8) and sdw is the DW of seaweed harvested annually per km2.Market value and climate benefits of seaweedFurther transportation and processing costs, economic value and net emissions of either sinking seaweed in the deep ocean for carbon sequestration or converting seaweed into usable products (biofuel, animal feed, pulses, vegetables, fruits, oil crops and cereals) were calculated on the basis of ranges of transport costs ($ tDW−1 km−1), transport emissions (tCO2-eq t−1 km−1), conversion cost ($ tDW−1), conversion emissions (tCO2-eq tDW−1), market value of product ($ tDW−1) and the emissions avoided by product (tCO2-eq tDW−1) in the literature (Table 1). Market value was treated as globally homogeneous and does not vary by region. Emissions avoided by products were determined by comparing estimated emissions related to seaweed production to emissions from non-seaweed products that could potentially be replaced by seaweed (including non-CO2 greenhouse gas emissions from land use)24. Other parameters used are distance to nearest port (km), water depth (m), spatially explicit sequestration fraction (%)57 and distance to optimal sinking location (km; cost-optimized for maximum emissions benefit considering transport emissions combined with spatially explicit sequestration fraction; see ‘Distance to sinking point calculation’ below). Each Monte Carlo simulation calculated the cost of both CDR via sinking seaweed and GHG emissions mitigation via seaweed products.For seaweed CDR, after the seaweed is harvested, it can either be sunk in the same location that it was grown, or be transported to a more economically favourable sinking location where more of the seaweed carbon would remain sequestered for 100 yr (see ‘Distance to optimal sinking point’ below). Immediately post-harvest, the seaweed still contains a large amount of water, requiring a conversion from dry mass to wet mass for subsequent calculations33:$$s_{ww} = frac{{s_{dw}}}{{0.1}}$$
(10)
where sww is the annual wet weight of seaweed harvested per km2 and sdw is the annual DW of seaweed harvested per km2.The cost to transport harvested seaweed to the optimal sinking location was calculated as:$$c_{swtsink} = c_{transbase} times d_{sink} times s_{ww}$$
(11)
where cswtsink is the total annual cost to transport harvested seaweed to the optimal sinking location, ctransbase is the cost to transport 1 ton of material 1 km on a barge, dsink is the distance in km to the economically optimized sinking location and sww is the annually harvested seaweed wet weight in t km−2 as in equation (10).The costs associated with transporting replacement equipment (for example, lines, buoys,anchors) to the farming location and hauling back used equipment at the end of its assumed lifetime (1 yr for seeded line, 5–20 yr for capital equipment by equipment type) in the sinking CDR pathway were calculated as:$$c_{eqtsink} = left( {c_{transbase} times left( {2 times d_{sink}} right) times m_{eq}} right) + (c_{transbase} times d_{port} times m_{eq})$$
(12)
where ceqtsink is the total annualized cost to transport both used and replacement equipment, ctransbase is the cost to transport 1 ton of material 1 km on a barge, meq is the annualized equipment mass in tons, dsink is the distance in km to the economically optimized sinking location and dport is the ocean distance to the nearest port in km. We assumed that the harvesting barge travels from the farming location directly to the optimal sinking location with harvested seaweed and replaced (used) equipment in tow (including used seeded line and annualized mass of used capital equipment), sinks the harvested seaweed, returns to the farm location and then returns to the nearest port (see Supplementary Fig. 16). These calculations assumed the shortest sea-route distance (see Distance to optimal sinking point).The total value of seaweed that is sunk for CDR was therefore calculated as:$$v_{sink} = frac{{left( {v_{cprice} – left( {c_{swtsink} + c_{eqtsink}} right)} right)}}{{s_{dw}}}$$
(13)
where vsink is the total value (cost, if negative) of seaweed farmed for CDR in $ tDW−1, vcprice is a theoretical carbon price, cswtsink is as calculated in equation (11), ceqtsink is as calculated in equation (12) and sdw is the annually harvested seaweed DW in t km−2. We did not assume any carbon price in our Monte Carlo simulations (vcprice is equal to zero), making vsink negative and thus representing a net cost.To calculate net carbon impacts, our model included uncertainty in the efficiency of using the growth and subsequent deep-sea deposition of seaweed as a CDR method. The uncertainty is expected to include the effects of reduced phytoplankton growth from nutrient competition, the relationship between air–sea gas exchange and overturning circulation (hereafter collectively referred to as the ‘atmospheric removal fraction’) and the fraction of deposited seaweed carbon that remains sequestered for at least 100 yr. The total amount of atmospheric CO2 removed by sinking seaweed was calculated as:$$e_{seqsink} = k_{atm} times k_{fseq} times frac{{tC}}{{tDW}} times frac{{tCO_2}}{{tC}}$$
(14)
where eseqsink is net atmospheric CO2 sequestered annually in t km−2, katm is the atmospheric removal fraction and kfseq is the spatially explicit fraction of sunk seaweed carbon that remains sequestered for at least 100 yr (see ref. 57).The emissions from transporting harvested seaweed to the optimal sinking location were calculated as:$$e_{swtsink} = e_{transbase} times d_{sink} times s_{ww}$$
(15)
where eswtsink is the total annual CO2 emissions from transporting harvested seaweed to the optimal sinking location in tCO2 km−2, etransbase is the CO2 emissions (tons) from transporting 1 ton of material 1 km on a barge (tCO2 per t-km), dsink is the distance in km to the economically optimized sinking location and sww is the annually harvested seaweed wet weight in t km−2 as in equation (10). Since the unit for etransbase is tCO2 per t-km, the emissions from transporting seaweed to the optimal sinking location are equal to (e_{mathrm{transbase}} times d_{mathrm{sink}} times s_{mathrm{ww}}), and the emissions from transporting seaweed from the optimal sinking location back to the farm are equal to 0 (since the seaweed has already been deposited, the seaweed mass to transport is now 0). Note that this does not yet include transport emissions from transport of equipment post-seaweed-deposition (see equation 16 below and Supplementary Fig. 16).The emissions associated with transporting replacement equipment (for example, lines, buoys, anchors) to the farming location and hauling back used equipment at the end of its assumed lifetime (1 yr for seeded line, 5–20 yr for capital equipment by equipment type)28,41 in the sinking CDR pathway were calculated as:$$e_{eqtsink} = left( {e_{transbase} times left( {2 times d_{sink}} right) times m_{eq}} right) + (e_{transbase} times d_{port} times m_{eq})$$
(16)
where eeqtsink is the total annualized CO2 emissions in tons from transporting both used and replacement equipment, etransbase is the CO2 emissions from transporting 1 ton of material 1 km on a barge, meq is the annualized equipment mass in tons, dsink is the distance in km to the economically optimized sinking location and dport is the ocean distance to the nearest port in km. We assumed that the harvesting barge travels from the farming location directly to the optimal sinking location with harvested seaweed and replaced (used) equipment in tow (including used seeded line and annualized mass of used capital equipment), sinks the harvested seaweed, returns to the farm location and then returns to the nearest port. These calculations assumed the shortest sea-route distance (see Distance to optimal sinking point).Net CO2 emissions removed from the atmosphere by sinking seaweed were thus calculated as:$$e_{remsink} = frac{{left( {e_{seqsink} – left( {e_{swtsink} + e_{eqtsink}} right)} right)}}{{s_{dw}}}$$
(17)
where eremsink is the net atmospheric CO2 removed per ton of seaweed DW, eseqsink is as calculated in equation (14), eswtsink is as calculated in equation (15), eeqtsink is as calculated in equation (16) and sdw is the annually harvested seaweed DW in t km−2.Net cost of climate benefitsSinkingTo calculate the total net cost and emissions from the production, harvesting and transport of seaweed for CDR, we combined the cost and emissions from the sinking-pathway cost and value modules. The total net cost of seaweed CDR per DW ton of seaweed was calculated as:$$c_{sinknet} = c_{prod} – v_{sink}$$
(18)
where csinknet is the total net cost of seaweed for CDR per DW ton harvested, cprod is the net production cost per DW ton as calculated in equation (6) and vsink is the net value (or cost, if negative) per ton seaweed DW as calculated in equation (13).The total net CO2 emissions removed per DW ton of seaweed were calculated as:$$e_{sinknet} = e_{remsink} – e_{prod}$$
(19)
where esinknet is the total net atmospheric CO2 removed per DW ton of seaweed harvested annually (tCO2 tDW−1 yr−1), eremsink is the net atmospheric CO2 removed via seaweed sinking annually as calculated in equation (17) and eprod is the net CO2 emitted from production and harvesting of seaweed annually as calculated in equation (9). For each Monte Carlo simulation, locations where esinknet is negative (that is, net emissions rather than net removal) were not included in subsequent calculations since they would not be contributing to CDR in that location under the given scenario. Note that these net emissions cases only occur in areas far from port in specific high-emissions scenarios. Even in such cases, most areas still contribute to CO2 removal (negative emissions), hence costs from locations with net removal were included.Total net cost was then divided by total net emissions to get a final value for cost per ton of atmospheric CO2 removed:$$c_{pertonsink} = frac{{c_{sinknet}}}{{e_{sinknet}}}$$
(20)
where cpertonsink is the total net cost per ton of atmospheric CO2 removed via seaweed sinking ($ per tCO2 removed), csinknet is total net cost per ton seaweed DW harvested as calculated in equation (18) ($ tDW−1) and esinknet is the total net atmospheric CO2 removed per ton seaweed DW harvested as calculated in equation (19) (tCO2 tDW−1).GHG emissions mitigationInstead of sinking seaweed for CDR, seaweed can be used to make products (including but not limited to food, animal feed and biofuels). Replacing convention products with seaweed-based products can result in ‘avoided emissions’ if the emissions from growing, harvesting, transporting and converting seaweed into products are less than the total greenhouse gas emissions (including non-CO2 GHGs) embodied in conventional products that seaweed-based products replace.When seaweed is used to make products, we assumed it is transported back to the nearest port immediately after being harvested. The annualized cost to transport the harvested seaweed and replacement equipment (for example, lines, buoys, anchors) was calculated as:$$c_{transprod} = frac{{left( {c_{transbase} times d_{port} times left( {s_{ww} + m_{eq}} right)} right)}}{{s_{dw}}}$$
(21)
where ctransprod is the annualized cost per ton seaweed DW to transport seaweed and equipment back to port from the farm location, ctransbase is the cost to transport 1 ton of material 1 km on a barge, meq is the annualized equipment mass in tons, dport is the ocean distance to the nearest port in km, sww is the annual wet weight of seaweed harvested per km2 as calculated in equation (10) and sdw is the annual DW of seaweed harvested per km2.The total value of seaweed that is used for seaweed-based products was calculated as:$$v_{product} = v_{mkt} – left( {c_{transprod} + c_{conv}} right)$$
(22)
where vproduct is the total value (cost, if negative) of seaweed used for products ($ tDW−1), vmkt is how much each ton of seaweed would sell for, given the current market price of conventional products that seaweed-based products replace ($ tDW−1), ctransprod is as calculated in equation (21) and cconv is the cost to convert each ton of seaweed to a usable product ($ tDW−1).The annualized CO2 emissions from transporting harvested seaweed and equipment back to port were calculated as:$$e_{transprod} = frac{{left( {e_{transbase} times d_{port} times left( {s_{ww} + m_{eq}} right)} right)}}{{s_{dw}}}$$
(23)
where etransprod is the annualized CO2 emissions per ton seaweed DW to transport seaweed and equipment back to port from the farm location, etransbase is the CO2 emissions from transporting 1 ton of material 1 km on a barge, meq is the annualized equipment mass in tons, dport is the ocean distance to the nearest port in km, sww is the annual wet weight of seaweed harvested per km2 as calculated in equation (10) and sdw is the annual DW of seaweed harvested per km2.Total emissions avoided by each ton of harvested seaweed DW were calculated as:$$e_{avprod} = e_{subprod} – left( {e_{transprod} + e_{conv}} right)$$
(24)
where eavprod is total CO2-eq emissions avoided per ton of seaweed DW per year (including non-CO2 GHGs using a GWP time period of 100 yr), esubprod is the annual CO2-eq emissions avoided per ton seaweed DW by replacing a conventional product with a seaweed-based product, etransprod is as calculated in equation (23) and econv is the annual CO2 emissions per ton seaweed DW from converting seaweed into usable products. esubprod was calculated by converting seaweed DW to caloric content58 for food/feed and comparing emissions intensity per kcal to agricultural products24, or by converting seaweed DW into equivalent biofuel content with a yield of 0.25 tons biofuel per ton DW59 and dividing the CO2 emissions per ton fossil fuel by the seaweed biofuel yield.To calculate the total net cost and emissions from the production, harvesting, transport and conversion of seaweed for products, we combined the cost and emissions from the product-pathway cost and value modules. The total net cost of seaweed for products per ton DW was calculated as:$$c_{prodnet} = c_{prod} – v_{product}$$
(25)
where cprodnet is the total net cost per ton DW of seaweed harvested for use in products, cprod is the net production cost per ton DW as calculated in equation (6) and vproduct is the net value (or cost, if negative) per ton DW as calculated in equation (22).The total net CO2-eq emissions avoided per ton DW of seaweed used in products were calculated as:$$e_{prodnet} = e_{avprod} – e_{prod}$$
(26)
where eprodnet is the total net CO2-eq emissions avoided per ton DW of seaweed harvested annually (tCO2 tDW−1 yr−1), eavprod is the net CO2-eq emissions avoided by seaweed products annually as calculated in equation (24) and eprod is the net CO2 emitted from production and harvesting of seaweed annually as calculated in equation (9). For each Monte Carlo simulation, locations where eprodnet is negative (that is, net emissions rather than net emissions avoided) were not included in subsequent calculations since they would not be avoiding any emissions in that scenario.Total net cost was then divided by total net emissions avoided to get a final value for cost per ton of CO2-eq emissions avoided:$$c_{pertonprod} = frac{{c_{prodnet}}}{{e_{prodnet}}}$$
(27)
where cpertonprod is the total net cost per ton of CO2-eq emissions avoided by seaweed products ($ per tCO2-eq avoided), cprodnet is total net cost per ton seaweed DW harvested for products as calculated in equation (25) ($ tDW−1) and eprodnet is total net CO2-eq emissions avoided per ton seaweed DW harvested for products as calculated in equation (26) (tCO2 tDW−1).Parameter ranges for Monte Carlo simulationsFor technoeconomic parameters with two or more literature values (see Supplementary Table 1), we assumed that the maximum literature value reflected the 95th percentile and the minimum literature value represented the 5th percentile of potential costs or emissions. For parameters with only one literature value, we added ±50% to the literature value to represent greater uncertainty within the modelled parameter range. Values at each end of parameter ranges were then rounded before Monte Carlo simulations as follows: capital costs, operating costs and harvest costs to the nearest $10,000 km−2, labour costs and insurance costs to the nearest $1,000 km−2, line costs to the nearest $0.05 m−1, transport costs to the nearest $0.05 t−1 km−1, transport emissions to the nearest 0.000005 tCO2 t−1 km−1, maintenance transport emissions to the nearest 0.0005 tCO2 km−1, product-avoided emissions to the nearest 0.1 tCO2-eq tDW−1, conversion cost down to the nearest $10 tDW−1 on the low end of the range and up to the nearest $10 tDW−1 on the high end of the range, and conversion emissions to the nearest 0.01 tCO2 tDW−1.We extended the minimum range values of capital costs to $10,000 km−2 and transport emissions to 0 to reflect potential future innovations, such as autonomous floating farm setups that would lower capital costs and net-zero emissions boats that would result in 0 transport emissions. To calculate the minimum value of $10,000 km−2 for a potential autonomous floating farm, we assumed that the bulk of capital costs for such a system would be from structural lines and flotation devices, and we therefore used the annualized structural line (system rope) and buoy costs from ref. 41 rounded down to the nearest $5,000 km−2. The full ranges used for our Monte Carlo simulations and associated literature values are shown in Supplementary Table 1.Distance to optimal sinking pointDistance to the optimal sinking point was calculated using a weighted distance transform (path-finding algorithm, modified from code in ref. 60) that finds the shortest ocean distance from each seaweed growth pixel to the location at which the net CO2 removed is maximized (including impacts of both increased sequestration fraction and transport emissions for different potential sinking locations) and the net cost is minimized. This is not necessarily the location in which the seaweed was grown, since the fraction of sunk carbon that remains sequestered for 100 yr is spatially heterogeneous (see ref. 57). For each ocean grid cell, we determined the cost-optimal sinking point by iteratively calculating equations (11–20) and assigning dsink as the distance calculated by weighted distance transform to each potential sequestration fraction (0.01–1.00) in increments of 0.01. Except for transport emissions, the economic parameter values used for these calculations were the averages of unrounded literature value ranges; we assumed that the maximum literature value reflected the 95th percentile and the minimum literature value represented the 5th percentile of potential costs or emissions, or for parameters with only one literature value, we added ±50% to the literature value to represent greater uncertainty within the modelled parameter range. For transport and maintenance transport emissions, we extended the minimum values of the literature ranges to zero to reflect potential net-zero emissions transport options and used the mean values of the resulting ranges. The dsink that resulted in minimum net cost per ton CO2 for each ocean grid cell was saved as the final dsink map, and the associated sequestration fraction value that the seaweed is transported to via dsink was assigned to the original cell where the seaweed was farmed and harvested (Supplementary Fig. 19). If the cost-optimal location to sink using this method was the same cell where the seaweed was harvested, then dsink was 0 km and the sequestration fraction was not modified from its original value (Supplementary Fig. 18).Comparison of gigaton-scale sequestration area to previous estimatesPrevious related work estimating the ocean area suitable for macroalgae cultivation13 and/or the area that might be required to reach gigaton-scale carbon removal via macroalgae cultivation13,19,36 has yielded a wide range of results, primarily due to differences in modelling methods. For example, Gao et al. (2022)36 estimate that 1.15 million km2 would be required to sequester 1 GtCO2 annually when considering carbon lost from seaweed biomass/sequestered as particulate organic carbon (POC) and refractory dissolved organic carbon (rDOC), and assume that the harvested seaweed is sold as food such that the carbon in the harvested seaweed is not sequestered. The area (0.31 million km2) required to sequester 1 GtCO2 in our study assumes that all harvested seaweed is sunk to the deep ocean to sequester carbon.Additionally, Wu et al.19 estimates that roughly 12 GtCO2 could be sequestered annually via macroalgae cultivation in approximately 20% of the world ocean area (that is, 1.67% ocean area per GtCO2), which is a much larger area per GtCO2 than our estimate of 0.085% ocean area. This notable difference arises for several reasons (including differences in yields, which in Wu et al. are around 500 tDW yr−1 in the highest-yield areas, whereas yields in our cheapest sequestration areas from G-MACMODS average 3,400 tDW km−2 yr−1) that arise from differences in model methodology. First, Wu et al. model temperate brown seaweeds, while our study considers different seaweed types, many of which have higher growth rates, and uses the most productive seaweed type for each ocean grid cell. The G-MACMODS seaweed growth model we use also has a highly optimized harvest schedule, includes luxury nutrient uptake (a key feature of macroalgal nutrient physiology) and does not directly model competition with phytoplankton during seaweed growth. Finally, tropical red seaweeds (the seaweed type in our cheapest sequestration areas) grow year-round, while others, such as the temperate brown seaweeds modelled by Wu et al., only grow seasonally. These differences all contribute to higher productivity in our model, leading to a smaller area required for gigaton-scale CO2 sequestration compared with Wu et al.Conversely, the ocean areas we model for seaweed-based CO2 sequestration or GHG emissions avoided are much larger than the 48 million km2 that Froehlich et al.13 estimate to be suitable for macroalgae farming globally. Although our maps show productivity and costs everywhere, the purpose of our modelling was to evaluate where different types of seaweed grow best and how production costs and product values vary over space, to highlight the lowest-cost areas (which are often the highest-producing areas) under various technoeconomic assumptions.Comparison of seaweed production costs to previous estimatesAlthough there are not many estimates of seaweed production costs in the scientific literature, our estimates for the lowest-cost 1% area of the ocean ($190–$2,790 tDW−1) are broadly consistent with previously published results: seaweed production costs reported in the literature range from $120 to $1,710 tDW−1 (refs. 40,41,61,62), but are highly dependent on assumed seaweed yields. For example, Camus et al.41 calculate a cost of $870 tDW−1 assuming a minimum yield of 12.4 kgDW m−1 of cultivation line (equivalent to 8.3 kgDW m−2 using 1.5 m spacing between lines). Using the economic values from Camus et al. but with our estimates of average yield for the cheapest 1% production cost areas (2.6 kgDW m−2) gives a much higher average cost of $2,730 tDW−1. Contrarily, van den Burg et al.40 calculate a cost of $1,710 tDW−1 using a yield of 20 tDW ha−1 (that is, 2.0 kg m−2). Instead assuming the average yield to be that from our lowest-cost areas (that is, 2.6 kgDW m−2 or 26 tDW ha−1) would decrease the cost estimated by van den Burg et al. (2016) to $1,290 tDW−1. Most recently, Capron et al.62 calculate an optimistic scenario cost of $120 tDW−1 on the basis of an estimated yield of 120 tDW ha−1 (12 kg m−2; over 4.5 times higher than the average yield in our lowest-cost areas). Again, instead assuming the average yield to be that in our lowest-cost areas would raise Capron et al.’s production cost to $540 tDW−1 (between the $190–$880 tDW−1 minimum to median production costs in the cheapest 1% areas from our model; Fig. 1a,b).Data sourcesSeaweed biomass harvestedWe used spatially explicit data for seaweed harvested globally under both ambient and limited-nutrient scenarios from the G-MACMODS seaweed growth model33.Fraction of deposited carbon sequestered for 100 yrWe used data from ref. 57 interpolated to our 1/12-degree grid resolution.Distance to the nearest portWe used the Distance from Port V1 dataset from Global Fishing Watch (https://globalfishingwatch.org/data-download/datasets/public-distance-from-port-v1) interpolated to our 1/12-degree grid resolution.Significant wave heightWe used data for annually averaged significant wave height from the European Center for Medium-range Weather Forecasts (ECMWF) interpolated to our 1/12-degree grid resolution.Ocean depthWe used data from the General Bathymetric Chart of the Oceans (GEBCO).Shipping lanesWe used data of Automatic Identification System (AIS) signal count per ocean grid cell, interpolated to our 1/12-degree grid resolution. We defined a major shipping lane grid cell as any cell with >2.25 × 108 AIS signals, a threshold that encompasses most major trans-Pacific and trans-Atlantic shipping lanes as well as major shipping lanes in the Indian Ocean, the North Sea, and coastal routes worldwide.Marine protected areas (MPAs)We used data from the World Database on Protected Areas (WDPA) and defined an MPA as any ocean MPA >20 km2.Reporting summaryFurther information on research design is available in the Nature Portfolio Reporting Summary linked to this article. More

Evans, N. P. et al. Quantification of drought during the collapse of the classic Maya civilization. Science 361, 498–501 (2018).Article 
ADS 
CAS 

Google Scholar 
Gill, R. B. The Great Maya Droughts: Water, Life, and Death (University of New Mexico Press, 2001).
Google Scholar 
Coe, M. D. The Maya (Thames and Hudson, 1993).
Google Scholar 
Douglas, P. M. J. et al. Drought, agricultural adaptation, and sociopolitical collapse in the Maya Lowlands. Proc. Natl. Acad. Sci. USA 112, 5607–5612 (2015).Article 
ADS 
CAS 

Google Scholar 
Haug, G. H. et al. Climate and the collapse of Maya civilization. Science 299, 1731–1735 (2003).Article 
ADS 
CAS 

Google Scholar 
Ford, A. & Nigh, R. Origins of the Maya forest garden: Maya resource management. J. Ethnobiol. 29, 213–236 (2009).Article 

Google Scholar 
Anderson, E. N. et al. Las Plantas de los Mayas: Etnobotánica en Quintana Roo, México (CONABIO-ECOSUR, 2005).
Google Scholar 
Fedick, S. L. Maya cornucopia: Indigenous food plants of the Maya lowlands. in The Real Business of Ancient Maya Economies (eds. Masson, M. A., Freidel, D. A. & Demarest, A. A.). 224–237 (University Press Florida, 2020).Ford, A. & Clarke, K. C. Linking the past and present of the ancient Maya: Lowland land use, population distribution, and density in the Late Classic period. in The Oxford Handbook of Historical Ecology and Applied Archaeology (eds. Isendahl, C. & Stump, D.) (Oxford Handbook of Historical Ecology and Applied Archaeology, 2015).Ford, A. & Nigh, R. The Maya Forest Garden: Eight Millennia of Sustainable Cultivation of the Tropical Woodlands (Routledge, 2016).Gómez-Pompa, A. On maya silviculture. Mexican Stud. (Estudios Mexicanos) 3, 1–17 (1987).Article 

Google Scholar 
Beach, T., Luzzadder-Beach, S., Krause, S. & Walling, S. ‘Mayacene’ floodplain and wetland formation in the Rio Bravo watershed of northwestern Belize. Holocene 25(10), 1612–1622 (2015).Pohl, M. D. et al. Early agriculture in the Maya lowlands. Lat. Am. Antiq. 7, 355–372 (1996).Article 

Google Scholar 
Fedick, S. L. The Managed Mosaic: Ancient Maya Agriculture and Resource Use (University of Utah Press, 1996).
Google Scholar 
Mueller, A. D. et al. Recovery of the forest ecosystem in the tropical lowlands of northern Guatemala after disintegration of Classic Maya polities. Geology 38, 523–526 (2010).Article 
ADS 

Google Scholar 
Hodell, D. A., Curtis, J. H. & Brenner, M. Possible role of climate in the collapse of Classic Maya civilization. Nature 375, 391–394 (1995).Article 
ADS 
CAS 

Google Scholar 
Islebe, G. A., Hooghiemstra, H., Brenner, M., Curtis, J. H. & Hodell, D. A. A Holocene vegetation history from lowland Guatemala. Holocene 6, 265–271 (1996).Article 
ADS 

Google Scholar 
Medina-Elizalde, M., Polanco-Martínez, J. M., Lases-Hernández, F., Bradley, R. & Burns, S. Testing the ‘tropical storm’ hypothesis of Yucatan Peninsula climate variability during the Maya Terminal Classic Period. Quat. Res. 86, 111–119 (2016).Aragón-Moreno, A. A., Islebe, G. A., Torrescano-Valle, N. & Arellano-Verdejo, J. Middle and late Holocene mangrove dynamics of the Yucatan Peninsula, Mexico. J. South Am. Earth Sci. 85, 307–311 (2018).Article 
ADS 

Google Scholar 
Aragón-Moreno, A. A., Islebe, G. A., Roy, P. D., Torrescano-Valle, N. & Mueller, A. D. Climate forcings on vegetation of the southeastern Yucatán Peninsula (Mexico) during the middle to late Holocene. Palaeogeogr. Palaeoclimatol. Palaeoecol. 495, 214–226 (2018).Article 

Google Scholar 
Kennett, D. J. et al. Development and disintegration of Maya political systems in response to climate change. Science 338, 788–791 (2012).Article 
ADS 
CAS 

Google Scholar 
Conde, C. et al. El Niño y la agricultura. in Los impactos de El Niño en México (ed. Magaña, V.). 103–135 (Dirección General de Protección Civil, Secretaría de Gobernación, México, 1999).Magaña, V. O., Vázquez, J. L., Pérez, J. L. & Pérez, J. B. Impact of El Niño on precipitation in Mexico. Geofísica Int. 42, 313–330 (2003).
Google Scholar 
Wahl, D., Byrne, R. & Anderson, L. An 8700 year paleoclimate reconstruction from the southern Maya lowlands. Quat. Sci. Rev. 103, 19–25 (2014).Article 
ADS 

Google Scholar 
Nooren, K. et al. Climate impact on the development of Pre-Classic Maya civilisation. Clim. Past 14, 1253–1273 (2018).Article 

Google Scholar 
Palomo-Kumul, J., Valdez-Hernández, M., Islebe, G. A., Cach-Pérez, M. J. & El Andrade, J. L. Niño-Southern oscillation affects the water relations of tree species in the Yucatan Peninsula. Mexico. Sci. Rep. 11, 10451 (2021).Article 
ADS 
CAS 

Google Scholar 
Rosenswig, R. M., VanDerwarker, A. M., Culleton, B. J. & Kennett, D. J. Is it agriculture yet? Intensified maize-use at 1000 cal BC in the Soconusco and Mesoamerica. J. Anthropol. Archaeol. 40, 89–108 (2015).Article 

Google Scholar 
Mueller, A. D. et al. Climate drying and associated forest decline in the lowlands of northern Guatemala during the late Holocene. Quat. Res. 71, 133–141 (2009).Article 

Google Scholar 
Aragón-Moreno, A. A., Islebe, G. A. & Torrescano-Valle, N. A ~3800-yr, high-resolution record of vegetation and climate change on the north coast of the Yucatan Peninsula. Rev. Palaeobot. Palynol. 178, 35–42 (2012).Article 

Google Scholar 
Carrillo-Bastos, A., Islebe, G. A. & Torrescano-Valle, N. 3800 Years of quantitative precipitation reconstruction from the Northwest Yucatan Peninsula. PLoS ONE 8, e84333 (2013).Article 
ADS 

Google Scholar 
Berglund, B. E. Human impact and climate changes—Synchronous events and a causal link?. Quat. Int. 105, 7–12 (2003).Article 

Google Scholar 
Vela-Peláez, A. A., Torrescano-Valle, N., Islebe, G. A., Mas, J. F. & Weissenberger, H. Holocene precipitation changes in the Maya forest, Yucatán peninsula. Mexico. Palaeogeogr. Palaeoclimatol. Palaeoecol. 505, 42–52 (2018).Article 
ADS 

Google Scholar 
Torrescano-Valle, N. & Islebe, G. A. Holocene paleoecology, climate history and human influence in the southwestern Yucatán Peninsula. Rev. Palaeobot. Palynol. 217, 1–8 (2015).Article 

Google Scholar 
Anselmetti, F. S., Hodell, D. A., Ariztegui, D., Brenner, M. & Rosenmeier, M. F. Quantification of soil erosion rates related to ancient Maya deforestation. Geology 35, 915–918 (2007).Article 
ADS 

Google Scholar 
Beach, T. et al. A review of human and natural changes in Maya Lowland wetlands over the Holocene. Quat. Sci. Rev. 28, 1710–1724 (2009).Article 
ADS 

Google Scholar 
Kerr, M. T. Holocene Precipitation Variability, Prehistoric Agriculture, and Natural and Human-Set Fires in Costa Rica (University of Tennessee, 2019).
Google Scholar 
Ebert, C. E., Peniche May, N., Culleton, B. J., Awe, J. J. & Kennett, D. J. Regional response to drought during the formation and decline of Preclassic Maya societies. Quat. Sci. Rev. 173, 211–235 (2017).Article 
ADS 

Google Scholar 
De la Barreda, B., Metcalfe, S. E. & Boyd, D. S. Precipitation regionalization, anomalies and drought occurrence in the Yucatan Peninsula, Mexico. Int. J. Climatol. 40, 4541–4555 (2020).Article 

Google Scholar 
Islebe, G. A. et al. Holocene Paleoecology and Paleoclimatology of south and south-eastern Mexico: A palynological approach. in Mexico´s Environmental Holocene and Anthropocene History (eds. Torrescano-Valle, N., Islebe, G. A. & Roy, P.) (Springer, 2019).Tuxill, J., Reyes, L. A., Moreno, L. L., Uicab, V. C. & Jarvis, D. I. All maize is not equal: Maize variety choices and Mayan foodways in rural Yucatan, Mexico. in Pre-Columbian Foodways: Interdisciplinary Approaches to Food, Culture, and Markets in Ancient Mesoamerica (eds. Staller, J. & Carrasco, M.) 467–486 (Springer, 2010).Torrescano-Valle, N., Ramírez-Barajas, P. J., Islebe, G. A., Vela-Pelaez, A. A. & Folan, W. J. Human influence versus natural climate variability. in The Holocene and Anthropocene Environmental History of Mexico: A Paleoecological Approach on Mesoamerica (eds. Torrescano-Valle, N., Islebe, G. A. & Roy, P. D.). 171–194 (Springer, 2019).Faegri, K. & Iversen, J. Textbook of Pollen Analysis (Wiley, 1989).
Google Scholar 
Ford, A. The Maya forest: A domesticated landscape. in The Maya World (eds. Hutson, S. R. & Ardren, T.). 519–539 (Routledge, 2020).Fedick, S. L. & Santiago, L. S. Large variation in availability of Maya food plant sources during ancient droughts. Proc. Natl. Acad. Sci. USA 119, 2115657118 (2022).Article 

Google Scholar 
Puleston, D. E. The role of ramón in Maya subsistence. in Maya Subsistence. 353–366 (Elsevier, 1982).Atran, S. et al. Itza Maya tropical agro-forestry [and comments and replies]. Curr. Anthropol. 34, 633–700 (1993).Article 

Google Scholar 
Dussol, L., Elliott, M., Michelet, D. & Nondédéo, P. Ancient Maya sylviculture of breadnut (Brosimum alicastrum Sw.) and sapodilla (Manilkara zapota (L.) P. Royen) at Naachtun (Guatemala): A reconstruction based on charcoal analysis. Quat. Int. 457, 29–42 (2017).Ebel, R., de Jesús Méndez Aguilar, M. & Putnam, H. R. Milpa: One sister got climate-sick. The impact of climate change on traditional Maya farming systems. Int. J. Sociol. Agric. Food (Online) 24, 175–199 (2018).
Google Scholar 
Hernández-González, O. & Vergara-Yoisura, S. Studies on the productivity of Brosimum alicastrum a tropical tree used for animal feed in the Yucatan Peninsula. Bothalia 22, 7 (2014).
Google Scholar 
Martínez-Ruiz, N. del R. & Larqué-Saavedra, A. Semilla de Ramón. in Alimentos Vegetales Autóctonos Iberoamericanos Subutilizados (eds. Sonia, S.-A. & Álvarez-Parrilla, E.). 177–192 (Fabro Editores, 2018).Hatfield, J. L. & Dold, C. Water-use efficiency: Advances and challenges in a changing climate. Front. Plant Sci. 10, 103 (2019).Article 

Google Scholar 
Basso, B. & Ritchie, J. T. Evapotranspiration in high-yielding maize and under increased vapor pressure deficit in the US Midwest. Agric. Environ. Lett. 3, 170039 (2018).Article 

Google Scholar 
Gregory, P. J., Simmonds, L. P. & Pilbeam, C. J. Soil type, climatic regime, and the response of water use efficiency to crop management. Agron. J. 92, 814–820 (2000).Article 

Google Scholar 
Moy, C. M., Seltzer, G. O., Rodbell, D. T. & Anderson, D. M. Variability of El Niño/Southern Oscillation activity at millennial timescales during the Holocene epoch. Nature 420, 162–165 (2002).Article 
ADS 
CAS 

Google Scholar 
Revelle, W. psych: Procedures for Psychological, Psychometric, and Personality Research. R package at https://CRAN.R-project.org/package=psych (2022).Wickham, H. & Bryan, J. readxl: Read Excel Files. R package at https://readxl.tidyverse.org/ (2022).Wei, T. et al. Package ‘corrplot’. Statistician 56, e24 (2017).
Google Scholar 
QGIS Development Team. QGIS Geographic Information System. QGIS Association at https://www.qgis.org (2022)Instituto Nacional de Estadistica Geographia e Informatica (INEGI). 1:1000000 Merida, Carta de Precipitacion. Merida, Yucatán, Mexico (1981). More

Odum, E. P. The strategy of ecosystem development. Science 164, 262–270 (1969).Article 
ADS 
CAS 

Google Scholar 
Gorham, E., Vitousek, P. M. & Reiners, W. A. The regulation of element budgets over the course of terrestrial ecosystem succession. Annu. Rev. Ecol. Syst. 10, 53–84 (1979).Article 
CAS 

Google Scholar 
Corman, J. R. et al. Foundations and frontiers of ecosystem science: Legacy of a classic paper (Odum 1969). Ecosystems 22, 1160–1172. https://doi.org/10.1007/s10021-018-0316-3 (2019).Article 

Google Scholar 
Santín, C. et al. Towards a global assessment of pyrogenic carbon from vegetation fires. Glob. Change Biol. 22, 76–91. https://doi.org/10.1111/gcb.12985 (2016).Article 
ADS 

Google Scholar 
Kominoski, J. S., Gaiser, E. E. & Baer, S. G. Advancing theories of ecosystem development through long-term ecological research. Bioscience 68, 554–562. https://doi.org/10.1093/biosci/biy070 (2018).Article 

Google Scholar 
Balch, J. K., Bradley, B. A., D’Antonio, C. M. & Gómez-Dans, J. Introduced annual grass increases regional fire activity across the arid western USA (1980–2009). Glob. Change Biol. 19, 173–183. https://doi.org/10.1111/gcb.12046 (2013).Article 
ADS 

Google Scholar 
Abatzoglou, J. T. & Kolden, C. A. Climate change in Western US deserts: Potential for increased wildfire and invasive annual grasses. Rangeland Ecol. Manag. 64(5), 471–478 (2011).Article 

Google Scholar 
Shi, H. et al. Historical cover trends in a sagebrush steppe ecosystem from 1985 to 2013: Links with climate, disturbance, and management. Ecosystems 21, 913–929. https://doi.org/10.1007/s10021-017-0191-3 (2018).Article 

Google Scholar 
Seyfried, M. S. & Wilcox, B. P. Scale and the nature of spatial variability: Field examples having implications for hydrologic modeling. Water Resour. Res. 31, 173–184. https://doi.org/10.1029/94WR02025 (1995).Article 
ADS 

Google Scholar 
Gasch, C. K., Huzurbazar, S. V. & Stahl, P. D. Description of vegetation and soil properties in sagebrush steppe following pipeline burial, reclamation, and recovery time. Geoderma 265, 19–26. https://doi.org/10.1016/j.geoderma.2015.11.013 (2016).Article 
ADS 

Google Scholar 
Huber, D. P. et al. Vegetation and precipitation shifts interact to alter organic and inorganic carbon storage in desert soils. Ecosphere 10, e02655. https://doi.org/10.1002/ecs2.2655 (2019).Article 

Google Scholar 
Knight, D. H., Jones, G. P., Reiners, W. A. & Romme, W. H. Mountains and Plains: The Ecology of Wyoming Landscapes (Yale University Press, 2014).
Google Scholar 
Patton, N. R., Lohse, K. A., Seyfried, M. S., Godsey, S. E. & Parsons, S. Topographic controls on soil organic carbon on soil mantled landscapes. Sci. Rep. 9, 6390. https://doi.org/10.1038/s41598-019-42556-5 (2019).Article 
ADS 
CAS 

Google Scholar 
Schwabedissen, S. G., Lohse, K. A., Reed, S. C., Aho, K. A. & Magnuson, T. S. Nitrogenase activity by biological soil crusts in cold sagebrush steppe ecosystems. Biogeochemistry 134, 57–76. https://doi.org/10.1007/s10533-017-0342-9 (2017).Article 
CAS 

Google Scholar 
You, Y. et al. Biological soil crust bacterial communities vary along climatic and shrub cover gradients within a sagebrush steppe ecosystem. Front. Microbiol. 12, 2365. https://doi.org/10.3389/fmicb.2021.569791 (2021).Article 

Google Scholar 
Burke, I. C., Reiners, W. A. & Olson, R. K. Topographic control of vegetation in a mountain big sagebrush steppe. Vegetation 84, 77–86 (1989).Article 

Google Scholar 
Poulos, M. J., Pierce, J. L., Flores, A. N. & Benner, S. G. Hillslope asymmetry maps reveal widespread, multi-scale organization. Geophys. Res. Lett. 39, 6. https://doi.org/10.1029/2012GL051283 (2012).Article 

Google Scholar 
Smith, T. & Bookhagen, B. Climatic and biotic controls on topographic asymmetry at the global scale. J. Geophys. Res.: Earth Surf. 126, e2020JF005692. https://doi.org/10.1029/2020JF005692Received22 (2021).Article 
ADS 

Google Scholar 
Seyfried, M., Link, T., Marks, D. & Murdock, M. Soil temperature variability in complex terrain measured using fiber-optic distributed temperature sensing. Vadose Zone J. 15, 6. https://doi.org/10.2136/vzj2015.09.0128 (2016).Article 

Google Scholar 
Chambers, J. C. et al. Resilience and resistance of sagebrush ecosystems: Implications for state and transition models and management treatments. Rangel. Ecol. Manage. 67, 440–454. https://doi.org/10.2111/REM-D-13-00074.1 (2014).Article 

Google Scholar 
Chambers, J. C. et al. Operationalizing resilience and resistance concepts to address invasive grass-fire cycles. Front. Ecol. Evol. 7, 2369. https://doi.org/10.3389/fevo.2019.00185 (2019).Article 

Google Scholar 
Boehm, A. R. et al. Slope and aspect effects on seedbed microclimate and germination timing of fall-planted seeds. Rangel. Ecol. Manage. 75, 58–67. https://doi.org/10.1016/j.rama.2020.12.003 (2021).Article 

Google Scholar 
Sankey, J. B., Germino, M. J., Sankey, T. T. & Hoover, A. N. Fire effects on the spatial patterning of soil properties in sagebrush steppe, USA: A meta-analysis. Int. J. Wildl. Fire 21, 545–556. https://doi.org/10.1071/WF11092 (2012).Article 

Google Scholar 
Fellows, A., Flerchinger, G., Seyfried, M. S. & Lohse, K. A. Rapid recovery of mesic mountain big sagebrush gross production and respiration following prescribed fire. Ecosystems 21, 1283–1294. https://doi.org/10.1007/s10021-017-0218-9 (2018).Article 

Google Scholar 
Vega, S. P. et al. Interaction of wind and cold-season hydrologic processes on erosion from complex topography following wildfire in sagebrush steppe. Earth Surf. Process. Landforms https://doi.org/10.1002/esp.4778 (2019).Article 

Google Scholar 
Xie, J., Li, Y., Zhai, C., Li, C. & Lan, Z. CO2 absorption by alkaline soils and its implication to the global carbon cycle. Environ. Geol. 56, 953–961 (2009).Article 
ADS 
CAS 

Google Scholar 
Stanbery, C., Pierce, J. L., Benner, S. G. & Lohse, K. On the rocks: Quantifying storage of inorganic soil carbon on gravels and determining pedon-scale variability. CATENA 157, 436–442. https://doi.org/10.1016/j.catena.2017.06.011 (2017).Article 
CAS 

Google Scholar 
Stanbery, C. et al. Controls on the presence and concentration of soil inorganic carbon in a semi-arid watershed. CATENA https://doi.org/10.2139/ssrn.4267018 (2023).Article 

Google Scholar 
Cerling, T. E. & Quade, J. Stable carbon and oxygen isotopes in soil carbonates. Geophys. Monogr. 78, 217–231 (1993).ADS 

Google Scholar 
Tappa, D. J., Kohn, M. J., McNamara, J. P., Benner, S. G. & Flores, A. N. Isotopic composition of precipitation in a topographically steep, seasonally snow-dominated watershed and implications of variations from the global meteoric water line. Hydrol. Process. 30, 4582–4592. https://doi.org/10.1002/hyp.10940 (2016).Article 
ADS 
CAS 

Google Scholar 
Salomons, W., Goudie, A. & Mook, W. G. Isotopic composition of calcrete deposits from Europe, Africa and India. Earth Surf. Process. 3, 43–57. https://doi.org/10.1002/esp.3290030105 (1978).Article 
CAS 

Google Scholar 
Salomons, W. & Mook, W. G. In Handbook of Environmental Isotope Geochemistry (eds P. Fritz & J. Fontes) Ch. 6, 241–269 (Elsevier, 1986).Bodí, M. B. et al. Wildland fire ash: Production, composition and eco-hydro-geomorphic effects. Earth Sci. Rev. 130, 103–127. https://doi.org/10.1016/j.earscirev.2013.12.007 (2014).Article 
ADS 
CAS 

Google Scholar 
Kéraval, B. et al. Soil carbon dioxide emissions controlled by an extracellular oxidative metabolism identifiable by its isotope signature. Biogeosciences 13, 6353–6362. https://doi.org/10.5194/bg-13-6353-2016 (2016).Article 
ADS 
CAS 

Google Scholar 
Goforth, B. R., Graham, R. C., Hubbert, K. R., Zanner, C. W. & Minnich, R. A. Spatial distribution and properties of ash and thermally altered soils after high-severity forest fire, southern California. Int. J. Wildland Fire 14, 343–354 (2005).Article 

Google Scholar 
Glossner, K. L. et al. Long-term suspended sediment and particulate organic carbon yields from the Reynolds Creek Experimental Watershed and Critical Zone Observatory. Hydrol. Process. 36, e14484. https://doi.org/10.1002/hyp.14484 (2022).Article 
CAS 

Google Scholar 
Seyfried, M. S. et al. Reynolds creek experimental watershed and critical zone observatory. Vadoze Zone J. 17, 180129. https://doi.org/10.2136/vzj2018.07.0129 (2018).Article 
CAS 

Google Scholar 
McIntyre, D. H. Cenozoic geology of the Reynolds Creek Experimental Watershed, Owyhee County, Idaho (Idaho Bureau of Mines and Geology, 1972).Earth Resources Observation and Science (EROS) Center, U. Image of the week: Burned Area Analysis for the Soda Fire, Idaho, https://eros.usgs.gov/media-gallery/image-of-the-week/burned-area-analysis-the-soda-fire-idaho (2015).Jenny, H. Factors of Soil Formation (McGraw-Hill, 1941).Book 

Google Scholar 
Kormos, P. R. et al. 31 years of hourly spatially distributed air temperature, humidity, and precipitation amount and phase from Reynolds Critical Zone Observatory. Earth Syst. Sci. Data 10, 1197–1205. https://doi.org/10.5194/essd-10-1197-2018 (2018).Article 
ADS 

Google Scholar 
Thomas, G. W. In Methods in Soil Analysis. Part 3. Chemical Methods (ed Sparks, D. L. ) (Soil Science Society of America and American Society of Agronomy, 1996).Brodie, C. R. et al. Evidence for bias in C and N concentrations and δ13C composition of terrestrial and aquatic organic materials due to pre-analysis acid preparation methods. Chem. Geol. 282, 67–83. https://doi.org/10.1016/j.chemgeo.2011.01.007 (2011).Article 
ADS 
CAS 

Google Scholar 
Patton, N. P., Lohse, K. A., Seyfried, M. S., Will, R. & Benner, S. G. Lithology and coarse fraction adjusted bulk density estimates for determining total organic carbon stocks in dryland soils. Geoderma 337, 844–852. https://doi.org/10.1016/j.geoderma.2018.10.036 (2019).Article 
ADS 
CAS 

Google Scholar 
McGuire, L. A., Rasmussen, C., Youberg, A. M., Sanderman, J. & Fenerty, B. Controls on the Spatial distribution of near-surface pyrogenic carbon on hillslopes 1 year following wildfire. J. Geophys. Res.: Earth Surf. 126, e2020JF005996. https://doi.org/10.1029/2020JF005996 (2021).Article 
ADS 

Google Scholar 
Jiménez-González, M. A. et al. Spatial distribution of pyrogenic carbon in Iberian topsoils estimated by chemometric analysis of infrared spectra. Sci. Total Env. 790, 148170. https://doi.org/10.1016/j.scitotenv.2021.148170 (2021).Article 
CAS 

Google Scholar 
Baldock, J. A. et al. Quantifying the allocation of soil organic carbon to biologically significant fractions. Soil Res. 51, 561–576. https://doi.org/10.1071/SR12374 (2013).Article 
CAS 

Google Scholar 
Sanderman, J. et al. Soil organic carbon fractions in the Great Plains of the United States: An application of mid-infrared spectroscopy. Biogeochemistry 156, 97–114. https://doi.org/10.1007/s10533-021-00755-1 (2021).Article 
CAS 

Google Scholar 
Sherrod, L. A., Dunn, G., Peterson, G. A. & Kolberg, R. L. Inorganic carbon analysis by modified pressure-calcimeter method. Soil Sci. Soc. Am. J. 66, 299–305 (2002).Article 
ADS 
CAS 

Google Scholar 
Mikutta, R., Kleber, M., Kaiser, K. & Jahn, R. Review. Soil Sci. Soc. Am. J. 69, 120–135. https://doi.org/10.2136/sssaj2005.0120 (2005).Article 
ADS 
CAS 

Google Scholar 
Risk, D., Nickerson, N., Creelman, C., McArthur, G. & Owens, J. Forced Diffusion soil flux: A new technique for continuous monitoring of soil gas efflux. Agric. For. Meteorol. 151, 1622–1631. https://doi.org/10.1016/j.agrformet.2011.06.020 (2011).Article 
ADS 

Google Scholar 
Fierer, N. & Schimel, J. P. Effects of drying–rewetting frequency on soil carbon and nitrogen transformations. Soil Biol. Biochem. 34, 777–787. https://doi.org/10.1016/S0038-0717(02)00007-X (2002).Article 
CAS 

Google Scholar 
Dane, J. H., Topp, G. C. & Campbell, G. S. In Methods of Soil Analysis: Physical Methods. Vol. 4 (ed SSSA) 721–738 (2002). More

Richard, E., Said, S., Hamann, J. L. & Gaillard, J. M. Daily, seasonal and annual variations in individual home range overlap of two sympatric spacies of deer. Can. J. Zool. 92, 853–859 (2014).Article 

Google Scholar 
Sorensen, A. A., Stenhouse, G. B., Bourbonnais, M. L. & Nelson, T. A. Effects of habitat quality and anthropogenic disturbance on grizzly bear (Ursus arctos horribilis) home-range fidelity. Can. J. Zool. 93, 857–865 (2015).Article 

Google Scholar 
van Beest, F. M., Rivrud, I. M., Loe, L. E., Milner, J. M. & Mysterud, A. What determines variation in home range size across spatiotemporal scales in a large browsing herbivore?. J. Anim. Ecol. 80, 771–785 (2011).Article 

Google Scholar 
Naidoo, R., Du, P., Weaver, G. S. L. C., Jago, M. & Wegmann, M. Factors affecting intraspecific variation in home range size of a large African herbivore. Landsc. Ecol. 27, 1523–1534 (2012).Article 

Google Scholar 
Bose, S. et al. Implications of fidelity and philopatry for the population structure of female black-tailed deer. Behav. Ecol. 28, 983–990 (2017).Article 

Google Scholar 
Northrup, J. M., Anderson, C. R. Jr. & Wittemyer, G. Environmental dynamics and anthropogenic development alter philopatry and space-use in a North American cervid. Divers. Distrib. 22, 547–557 (2016).Article 

Google Scholar 
Passadore, C., Möller, L., Diaz-aguirre, F. & Parra, G. J. High site fidelity and restricted ranging patterns in southern Australian bottlenose dolphins. Ecol. Evol. 8, 242–256 (2018).Article 

Google Scholar 
Morales, J. M. et al. Building the bridge between animal movement and population dynamics. Philos. Trans. R. Soc. B Biol. Sci. 365, 2289–2301 (2010).Article 

Google Scholar 
Shaw, A. K. Causes and consequences of individual variation in animal movement. Mov. Ecol. 8, 1–12 (2020).Article 

Google Scholar 
Morrison, T. A. et al. Drivers of site fidelity in ungulates. J. Anim. Ecol. 00, 1–12 (2021).
Google Scholar 
Abrahms, B. et al. Emerging perspectives on resource tracking and animal movement ecology. Trends Ecol. Evol. 36, 308–320 (2021).Article 

Google Scholar 
Barraquand, F. & Benhamou, S. Animal movements in heterogeneous landscapes: Identifying profitable places and homogeneous movement bouts. Ecology 89, 3336–3348 (2008).Article 

Google Scholar 
Mueller, T. & Fagan, W. F. Search and navigation in dynamic environments: From individual behaviors to population distributions. Oikos 117, 654–664 (2008).Article 

Google Scholar 
Sawyer, H., Merkle, J. A., Middleton, A. D., Dwinnell, S. P. H. & Monteith, K. L. Migratory plasticity is not ubiquitous among large herbivores. J. Anim. Ecol. 88, 450–460 (2019).
Google Scholar 
Shakeri, Y. N., White, K. S. & Waite, J. N. Staying close to home: Ecological constraints on space use and range fidelity in a mountain ungulate. Ecol. Evol. 11, 11051–11064 (2021).Article 

Google Scholar 
Damuth, J. Home range, home range overlap, and species energy use among herbivorous mammals. Biol. J. Linn. Soc. 15, 185–193 (1981).Article 

Google Scholar 
Lindstedt, S. L., Miller, B. J. & Buskirk, S. W. Home range, time, and body size in mammals. Ecol. Soc. Am. 67, 413–418 (1986).
Google Scholar 
Ofstad, E. G., Herfindal, I., Solberg, E. J. & Sæther, B. E. Home ranges, habitat and body mass: Simple correlates of home range size in ungulates. Proc. R. Soc. B Biol. Sci. 283, 20161234 (2016).Article 

Google Scholar 
Gehr, B. et al. Stay home, stay safe—Site familiarity reduces predation risk in a large herbivore in two contrasting study sites. J. Anim. Ecol. 89, 1329–1339 (2020).Article 

Google Scholar 
Sach, F., Dierenfeld, E. S., Langley-Evans, S. C., Watts, M. J. & Yon, L. African savanna elephants (Loxodonta africana) as an example of a herbivore making movement choices based on nutritional needs. PeerJ 7, 1–27 (2019).Article 

Google Scholar 
Pretorius, Y. et al. Diet selection of African elephant over time shows changing optimization currency. Oikos 121, 2110–2120 (2012).Article 

Google Scholar 
Chamaillé-Jammes, S., Valeix, M. & Fritz, H. Managing heterogeneity in elephant distribution: Interactions between elephant population density and surface-water availability. J. Appl. Ecol. 44, 625–633 (2007).Article 

Google Scholar 
Purdon, A. & van Aarde, R. J. Water provisioning in Kruger National Park alters elephant spatial utilisation patterns. J. Arid Environ. 141, 45–51 (2017).Article 
ADS 

Google Scholar 
Shannon, G., Matthews, W. S., Page, B. R., Parker, G. E. & Smith, R. J. The affects of artificial water availability on large herbivore ranging patterns in savanna habitats: A new approach based on modelling elephant path distributions. Divers. Distrib. 15, 776–783 (2009).Article 

Google Scholar 
Kos, M. et al. Seasonal diet changes in elephant and impala in mopane woodland. Eur. J. Wildl. Res. 58, 279–287 (2012).Article 

Google Scholar 
Shannon, G., Mackey, R. L. & Slotow, R. Diet selection and seasonal dietary switch of a large sexually dimorphic herbivore. Acta Oecologica 46, 48–55 (2013).Article 
ADS 

Google Scholar 
Loarie, S. R., van Aarde, R. J. & Pimm, S. L. Elephant seasonal vegetation preferences across dry and wet savannas. Biol. Conserv. 142, 3099–3107 (2009).Article 

Google Scholar 
Scogings, P. F. et al. Seasonal variations in nutrients and secondary metabolites in semi-arid savannas depend on year and species. J. Arid Environ. 114, 54–61 (2015).Article 
ADS 

Google Scholar 
Birkett, P. J., Vanak, A. T., Muggeo, V. M. R., Ferreira, S. M. & Slotow, R. Animal perception of seasonal thresholds: Changes in elephant movement in relation to rainfall patterns. PLoS ONE 7, 1–8 (2012).Article 

Google Scholar 
Cushman, S. A., Chase, M. & Griffin, C. Elephants in space and time. Oikos 109, 331–341 (2005).Article 

Google Scholar 
Bohrer, G., Beck, P. S., Ngene, S. M., Skidmore, A. K. & Douglas-Hamilton, I. Elephant movement closely tracks precipitation-driven vegetation dynamics in a Kenyan forest-savanna landscape. Mov. Ecol. 2, 1–12 (2014).Article 

Google Scholar 
Purdon, A., Mole, M. A., Chase, M. J. & van Aarde, R. J. Partial migration in savanna elephant populations distributed across southern Africa. Sci. Rep. 8, 1–11 (2018).Article 
CAS 

Google Scholar 
Shannon, G., Page, B. R., Duffy, K. J. & Slotow, R. The ranging behaviour of a large sexually dimorphic herbivore in response to seasonal and annual environmental variation. Austral Ecol. 35, 731–742 (2010).Article 

Google Scholar 
Tsalyuk, M., Kilian, W., Reineking, B. & Getz, W. M. Temporal variation in resource selection of African elephants follows long-term variability in resource availability. Ecol. Monogr. 89, 1–19 (2019).Article 

Google Scholar 
Thaker, M., Prins, H. H. T., Slotow, R., Vanak, A. T. & Gupte, P. R. Fine-scale tracking of ambient temperature and movement reveals shuttling behavior of elephants to water. Front. Ecol. Evol. 7, 1–12 (2019).Article 

Google Scholar 
Govender, N., Trollope, W. S. W. & Van Wilgen, B. W. The effect of fire season, fire frequency, rainfall and management on fire intensity in savanna vegetation in South Africa. J. Appl. Ecol. 43, 748–758 (2006).Article 

Google Scholar 
MacFadyen, S., Hui, C., Verburg, P. H. & Van Teeffelen, A. J. A. Spatiotemporal distribution dynamics of elephants in response to density, rainfall, rivers and fire in Kruger National Park, South Africa. Divers. Distrib. 25, 880–894 (2019).Article 

Google Scholar 
Edwards, M. A., Nagy, J. A. & Derocher, A. E. Low site fidelity and home range drift in a wide-ranging, large Arctic omnivore. Anim. Behav. 77, 23–28 (2009).Article 

Google Scholar 
Switzer, P. Site fidelity in predictable and unpredictable habitats. Evol. Ecol. 7, 533–555 (1993).Article 

Google Scholar 
Kranstauber, B., Kays, R., Lapoint, S. D., Wikelski, M. & Safi, K. A dynamic Brownian bridge movement model to estimate utilization distributions for heterogeneous animal movement. J. Anim. Ecol. 81, 738–746 (2012).Article 

Google Scholar 
Kranstauber, B., Smolla, M. & Safi, K. Similarity in spatial utilization distributions measured by the earth mover’s distance. Methods Ecol. Evol. 8, 155–160 (2017).Article 

Google Scholar 
Wartmann, F., Juarez, C. & Fernandez-duque, E. Size, site fidelity, and overlap of home ranges and core areas in the socially monogamous owl monkey (Aotus azarae) of Northern Argentina. Int. J. Primatol. 35, 919–939 (2014).Article 

Google Scholar 
Pringle, R. M. Elephants as agents of habitat creation for small vertebrates at the patch scale. Ecology 89, 26–33 (2008).Article 

Google Scholar 
Valeix, M. et al. Elephant-induced structural changes in the vegetation and habitat selection by large herbivores in an African savanna. Biol. Conserv. 144, 902–912 (2011).Article 

Google Scholar 
Coverdale, T. C. et al. Elephants in the understory: opposing direct and indirect effects of consumption and ecosystem engineering by megaherbivores. Ecology 97, 3219–3230 (2016).Article 

Google Scholar 
Gertenbach, W. Rainfall patterns in the Kruger National Park. Koedoe 23, 35–43 (1980).Article 

Google Scholar 
Venter, F. J., Scholes, R. J. & Eckhardt, H. C. The abiotic template and its associated vegetation pattern. In The Kruger Experience (eds du Toit, J. T. et al.) 83–129 (Island Press, 2003).
Google Scholar 
Young, K. D., Ferreira, S. M. & van Aarde, R. J. The influence of increasing population size and vegetation productivity on elephant distribution in the Kruger National Park. Austral Ecol. 34, 329–342 (2009).Article 

Google Scholar 
Ferreira, S. M., Greaver, C. & Simms, C. Elephant population growth in Kruger National Park, South Africa, under a landscape management approach. Koedoe 59, 1–6 (2017).Article 

Google Scholar 
Brownrigg, R. Package ‘Maps’: Draw Geographical Maps (2022).Kranstauber, B. & Smolla, M. Move: Visualizing and analyzing animal track data. R package version 2.1.0 (2013).R Core Team. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing. URL https://www.R-project.org/ (2017).Horne, J. S., Garton, E. O., Krone, S. M. & Lewis, J. S. Analyzing animal movement using Brownian bridges. Ecology 88, 2354–2363 (2007).Article 

Google Scholar 
Wato, Y. A. et al. Movement patterns of African elephants (Loxodonta africana) in a semi-arid savanna suggest that they have information on the location of dispersed water sources. Front. Ecol. Evol. 6, 1–8 (2018).Article 

Google Scholar 
Polansky, L., Kilian, W. & Wittemyer, G. Elucidating the significance of spatial memory on movement decisions by African savannah elephants using state-space models. Proc. R. Soc. B Biol. Sci. 282, 1–7 (2015).
Google Scholar 
Archibald, S. & Scholes, R. J. Leaf green-up in a semi-arid African savanna–separating tree and grass responses to environmental cues. J. Veg. Sci. 18, 583–594 (2007).
Google Scholar 
Majozi, N. P. et al. Analysing surface energy balance closure and partitioning over a semi-arid savanna FLUXNET site in Skukuza, Kruger National Park, South Africa. Hydrol. Earth Syst. Sci. 21, 3401–3415 (2017).Article 
ADS 
CAS 

Google Scholar 
Dodge, S. et al. The environmental-data automated track annotation (Env-DATA) system: Linking animal tracks with environmental data. Mov. Ecol. 1, 1–14 (2013).Article 

Google Scholar 
Didan, K. MOD13Q1 MODIS/terra vegetation indices 16-day L3 global 250m SIN Grid V006. NASA EOSDIS Land Process. DAAC https://doi.org/10.5067/MODIS/MOD13Q1.006 (2015).Redfern, J. V., Grant, C. C., Gaylard, A. & Getz, W. M. Surface water availability and the management of herbivore distributions in an African savanna ecosystem. J. Arid Environ. 63, 406–424 (2005).Article 
ADS 

Google Scholar 
Young, K. D., Ferreira, S. M. & van Aarde, R. J. Elephant spatial use in wet and dry savannas of southern Africa. J. Zool. 278, 189–205 (2009).Article 

Google Scholar 
Goldenberg, S. Z., Douglas-Hamilton, I. & Wittemyer, G. Inter-generational change in African elephant range use is associated with poaching risk, primary productivity and adult mortality. Proc. R. Soc. B Biol. Sci. 285, 1–8 (2018).
Google Scholar 
Woolley, L.-A. et al. Population and individual elephant response to a catastrophic fire in Pilanesberg National Park. PLoS ONE 3, 1–10 (2008).Article 

Google Scholar 
Eby, S. L., Anderson, T. M., Mayemba, E. P. & Ritchie, M. E. The effect of fire on habitat selection of mammalian herbivores: The role of body size and vegetation characteristics. J. Anim. Ecol. 83, 1196–1205 (2014).Article 

Google Scholar 
Brooks, M. E. et al. glmmTMB balances speed and flexibility among packages for zero-inflated generalized linear mixed modeling. R J. 9, 378–400 (2017).Article 

Google Scholar 
Burnham, K. P. & Anderson, D. R. Model Selection and Multimodal Inference: A Practical Information-Theoretic Approach (Springer, 2002).MATH 

Google Scholar 
Mazerolle, M. J. AICcmodavg: Model Selection and Multimodel Inference Based on (Q)AIC(c) (2020).van Moorter, B. et al. Memory keeps you at home: A mechanistic model for home range emergence. Oikos 118, 641–652 (2009).Article 

Google Scholar 
Guldemond, R. A. R., Purdon, A. & van Aarde, R. J. A systematic review of elephant impact across Africa. PLoS ONE 12, 1–12 (2017).Article 

Google Scholar 
Abraham, J. O., Goldberg, E. R., Botha, J. & Staver, A. C. Heterogeneity in African savanna elephant distributions and their impacts on trees in Kruger National Park, South Africa. Ecol. Evol. 11, 5624–5634 (2021).Article 

Google Scholar 
Wall, J., Douglas-Hamilton, I. & Vollrath, F. Elephants avoid costly mountaineering. Curr. Biol. 16, 527–529 (2006).Article 

Google Scholar 
Presotto, A., Fayrer-Hosken, R., Curry, C. & Madden, M. Spatial mapping shows that some African elephants use cognitive maps to navigate the core but not the periphery of their home ranges. Anim. Cogn. 22, 251–263 (2019).Article 

Google Scholar 
Landman, M., Schoeman, D. S., Hall-Martin, A. J. & Kerley, G. I. H. Understanding long-term variations in an elephant piosphere effect to manage impacts. PLoS ONE 7, 1–11 (2012).Article 

Google Scholar 
Fahrig, L. et al. Functional landscape heterogeneity and animal biodiversity in agricultural landscapes. Ecol. Lett. 14, 101–112 (2011).Article 

Google Scholar 
Hamm, M. & Drossel, B. Habitat heterogeneity hypothesis and edge effects in model metacommunities. J. Theor. Biol. 426, 40–48 (2017).Article 
ADS 

Google Scholar 
Katayama, N. et al. Landscape heterogeneity-biodiversity relationship: Effect of range size. PLoS ONE 9, 1–8 (2014).Article 

Google Scholar 
Tews, J. et al. Animal species diversity driven by habitat heterogeneity/diversity: The importance of keystone structures. J. Biogeogr. 31, 79–92 (2004).Article 

Google Scholar 
O’Connor, T. G., Goodman, P. S. & Clegg, B. A functional hypothesis of the threat of local extirpation of woody plant species by elephant in Africa. Biol. Conserv. 136, 329–345 (2007).Article 

Google Scholar 
Codron, J. et al. Elephant (Loxodonta africana) diets in Kruger National Park, South Africa: Spatial and landscape differences. J. Mammal. 87, 27–34 (2006).Article 

Google Scholar 
Mduma, S. A. R., Sinclair, A. R. E. & Hilborn, R. Food regulates the Serengeti wildebeest: A 40-year record. J. Anim. Ecol. 68, 1101–1122 (1999).Article 

Google Scholar 
Ogutu, J. O. & Owen-Smith, N. ENSO, rainfall and temperature influences on extreme population declines among African savanna ungulates. Ecol. Lett. 6, 412–419 (2003).Article 

Google Scholar 
Codron, J. et al. Landscape-scale feeding patterns of African elephant inferred from carbon isotope analysis of feces. Oecologia 165, 89–99 (2011).Article 
ADS 

Google Scholar 
Woolley, L.-A., Millspaugh, J. J., Woods, R. J., Page, B. R. & Slotow, R. Intraspecific strategic responses of African elephants to temporal variation in forage quality. J. Wildl. Manag. 73, 827–835 (2009).Article 

Google Scholar 
Dube, K. & Nhamo, G. Evidence and impact of climate change on South African national parks. Potential implications for tourism in the Kruger National Park. Environ. Dev. 33, 1–11 (2020).Article 

Google Scholar 
Tshipa, A. et al. Partial migration links local surface-water management to large-scale elephant conservation in the world’s largest transfrontier conservation area. Biol. Conserv. 215, 46–50 (2017).Article 

Google Scholar 
Nathan, R. et al. Big-data approaches lead to an increased understanding of the ecology of animal movement. Science (80-.) 375, 1–12 (2022).Article 

Google Scholar 
Kays, R., Crofoot, M. C., Jetz, W. & Wikelski, M. Terrestrial animal tracking as an eye on life and planet. Science (80-.) 348, 1222–1232 (2015).Article 
CAS 

Google Scholar 
Mpakairi, K. S., Ndaimani, H., Tagwireyi, P., Zvidzai, M. & Madiri, T. H. Futuristic climate change scenario predicts a shrinking habitat for the African elephant (Loxodonta africana): Evidence from Hwange National Park, Zimbabwe. Eur. J. Wildl. Res. 66, 1–10 (2020).Article 

Google Scholar 
Staver, A. C., Wigley-Coetsee, C. & Botha, J. Grazer movements exacerbate grass declines during drought in an African savanna. J. Ecol. 107, 1482–1491 (2019).Article 

Google Scholar 
Asner, G. P., Vaughn, N., Smit, I. P. J. & Levick, S. Ecosystem-scale effects of megafauna in African savannas. Ecography (Cop.) 39, 240–252 (2016).Article 

Google Scholar 
Shannon, G. et al. Relative impacts of elephant and fire on large trees in a savanna ecosystem. Ecosystems 14, 1372–1381 (2011).Article 

Google Scholar 
Mole, M. A., DÁraujo, S. R., van Aarde, R. J., Mitchell, D. & Fuller, A. Coping with heat: Behavioural and physiological responses of savanna elephants in their natural habitat. Conserv. Physiol. 4, 1–11 (2016).Article 

Google Scholar 
Ncongwane, K. P., Botai, J. O., Sivakumar, V., Botai, C. M. & Adeola, A. M. Characteristics and long-term trends of heat stress for South Africa. Sustainability 13, 1–20 (2021).Article 

Google Scholar 
Lagendijk, G., Mackey, R. L., Page, B. R. & Slotow, R. The effects of herbivory by a mega- and mesoherbivore on tree recruitment in sand forest, South Africa. PLoS ONE 6, 1–9 (2011).Article 

Google Scholar 
Wells, H. B. M. et al. Experimental evidence that effects of megaherbivores on mesoherbivore space use are influenced by species’ traits. J. Anim. Ecol. 90, 2510–2522 (2021).Article 

Google Scholar 
Thaker, M. et al. Minimizing predation risk in a landscape of multiple predators: Effects on the spatial distribution of African ungulates. Ecology 92, 398–407 (2011).Article 

Google Scholar 
Fležar, U. et al. Simulated elephant-induced habitat changes can create dynamic landscapes of fear. Biol. Conserv. 237, 267–279 (2019).Article 

Google Scholar 
Brennan, A. et al. Characterizing multispecies connectivity across a transfrontier conservation landscape. J. Appl. Ecol. 57, 1700–1710 (2020).Article 

Google Scholar 
Roever, C. L., van Aarde, R. J. & Leggett, K. Functional connectivity within conservation networks: Delineating corridors for African elephants. Biol. Conserv. 157, 128–135 (2013).Article 

Google Scholar 
Green, S. E., Davidson, Z., Kaaria, T. & Doncaster, C. P. Do wildlife corridors link or extend habitat? Insights from elephant use of a Kenyan wildlife corridor. Afr. J. Ecol. 56, 860–871 (2018).Article 

Google Scholar  More