Philippa Kaur

Bowman, D., Williamson, G. J., Gibson, R. K., Bradstock, R. A. & Keenan, R. J. The severity and extent of the Australia 2019–20 Eucalyptus forest fires are not the legacy of forest management. Nat. Ecol. Evol. 5, 1003–1010 (2021).Article 

Google Scholar 
Lindenmayer, D. B., Kooyman, R., Taylor, C., Ward, M. & Watson, J. Recent Australian wildfires made worse by logging and associated forest management. Nat. Ecol. Evol. 4, 898–900 (2020).Article 

Google Scholar 
Gould, J. S., Knight, I. & Sullivan, A. L. Physical modelling of leaf scorch height from prescribed fires in young Eucalyptus sieberi regrowth forests in South-Eastern Australia. Int. J. Wildl. Fire 7, 7–20 (1997).Article 

Google Scholar 
Keith, D. Ocean Shores to Desert Dunes: The Native Vegetation of NSW and the ACT (Department of Environment and Conservation NSW, 2004).Burrows, N. Predicting canopy scorch height in jarrah forests. CALM Sci. 2, 267–274 (1997).
Google Scholar 
Penney, G., Habibi, D. & Cattani, M. Firefighter tenability and its influence on wildfire suppression. Fire Saf. J. 106, 38–51 (2019).Article 

Google Scholar 
Sharples, J. J. et al. Natural hazards in Australia: extreme bushfire. Clim. Change 139, 85–99 (2016).Article 

Google Scholar 
Attiwill, P. M. et al. Timber harvesting does not increase fire risk and severity in wet eucalypt forests of Southern Australia. Conserv. Lett. 7, 341–354 (2014).Article 

Google Scholar 
Lindenmayer, D., Taylor, C. & Blanchard, W. Empirical analyses of the factors influencing fire severity in southeastern Australia. Ecosphere 12, e03721 (2021).Article 

Google Scholar 
Taylor, C., Blanchard, W. & Lindenmayer, D. B. Does forest thinning reduce fire severity in Australian eucalypt forests? Conserv. Lett. 14, e12766 (2020).
Google Scholar 
Taylor, C., McCarthy, M. A. & Lindenmayer, D. B. Non-linear effects of stand age on fire severity. Conserv. Lett. 7, 355–370 (2014).Article 

Google Scholar 
Furlaud, J. M., Prior, L. D., Williamson, G. J. & Bowman, D. M. J. S. Fire risk and severity decline with stand development in Tasmanian giant Eucalyptus forest. For. Ecol. Manag. 502, 119724 (2021).Article 

Google Scholar 
Price, O. F. & Bradstock, R. A. The efficacy of fuel treatment in mitigating property loss during wildfires: insights from analysis of the severity of the catastrophic fires in 2009 in Victoria, Australia. J. Environ. Manag. 113, 146–157 (2012).Article 

Google Scholar 
Taylor, C. & Lindenmayer, D. B. The adequacy of Victoria’s protected areas for conserving its forest-dependent fauna. Austral Ecol. 44, 1076–1090 (2019).Article 

Google Scholar 
Taylor, C., Blanchard, W. & Lindenmayer, D. B. What are the relationships between thinning and fire severity? Austral Ecol. https://doi.org/10.1111/aec.13096 (2021).La Sala, A. Thinning Regrowth Eucalypts Native Forest Silviculture Technical Bulletin No. 13 (Forestry Tasmania, 2001).Cary, G. J., Blanchard, W., Foster, C. N. & Lindenmayer, D. B. Effects of altered fire intervals on critical timber production and conservation values. Int. J. Wildl. Fire 30, 322–328 (2021).Article 

Google Scholar 
Filkov, A. I. et al. The determinants of crown fire runs during extreme wildfires in broadleaf forests in Australia. Adv. For. Fire Res. https://doi.org/10.14195/978-989-26-16-506_190; http://hdl.handle.net/10316.2/44517 (2018).Lindenmayer, D. B., Hobbs, R. J., Likens, G. E., Krebs, C. J. & Banks, S. C. Newly discovered landscape traps produce regime shifts in wet forests. Proc. Natl Acad. Sci. USA 108, 15887–15891 (2011).CAS 
Article 

Google Scholar  More

Species used in the experimentsMilnesium inceptum32 (Fig. 1A, a picture taken using Olympus BX41 Phase Contrast light Microscope associated with Olympus SC50 digital camera) is an obligatory predatory species with the body length ranging from 326 to 848 μm. It feeds on rotifers, nematodes and other tardigrades and lays smooth eggs in exuviae. To stay active, M. inceptum needs a thin water film around its body14. The species inhabits places exposed to shorter and longer periods of drying i.e. frequently drying mosses growing on cement walls32. Till now it was reported in Poland, Germany, Japan, Switzerland and Bulgaria32. At the same time, it is a perfect organism for our research because (1) it is large and easy to observe, (2) it tolerates frequent periods of entering and leaving anhydrobiosis, (3) it easily creates a tun stage. Milnesium inceptum for experimental purposes were acquired from a moss sample from a cement wall in Poznań, Poland (52°24′15″N, 16°53′18″E). The extraction of tardigrades was conducted under stereomicroscope (Olympus SZ51) using standard methods33. Then specimens, further used in our experiments, have been cultured based on protocol proposed by Roszkowska et al.34. Only fully active, adult specimens were selected for the experiments.Figure 1Model animals used in experiments: (A) Milnesium inceptum; insert shows tardigrade in the tun state; (B) Cepaea nemoralis in its natural environment; (C) a tardigrade that appeared on moss surface during in vivo observation of rehydrated moss cushion (red arrow). Figures were assembled in Corel Photo-Paint 2017 (http://www.corel.com).Full size imageCepaea nemoralis35 (Fig. 1B, a picture taken using Motorola g(9), Camera version 7.3.63.53-whitney) is a stylommatophoran European land snail species, which is widespread and common throughout the continent36. The average maximum shell diameter is 20 to 22 mm37. It feeds on plant materials available, yet has a strong preference for dead and senescent herbs38. C. nemoralis occurs in variable habitats (frequently in synanthropic ones) such as forests, meadows, gardens, near shrubs or dunes36.The period of its activity falls on the growing season; it usually comes out of the shell and crawls when the air humidity reaches 70% or more, independently from solar radiation and air temperature28. The species is a good model for our study due to its: (1) large size compared to tardigrades, and (2) co-occurrence with M. inceptum in natural environments. Individuals of C. nemoralis were harvested from anthropogenic environment: gardens adjacent to detached houses (52°25′28″N, 16°46′52″E). Snails were collected from plants, cement walls and ground surfaces. After collection, all C. nemoralis specimens were washed-up and placed in 30 L (480 × 360 × 252 mm) transparent plastic box with mesh covering for ventilation. Soil and rocks were placed in the box allowing to maintain a moist shelter for snails, and a sepia was used as a source of a calcium. Animals were fed with lettuce, cabbage and nettle twice a week and sprinkled with water to stimulate their activity. Box containing snails was kept in a rearing room, at 17 °C in 12:12 photoperiod. Snails were kept in the box for 1.5 months prior to the experiments. For the experiments we used only adult animals. The snails were checked under Olympus SZX7 stereomicroscope prior to the experiment to ensure they were free of tardigrades.Pilot studiesDoes the tardigrades’ distribution within a moss cushion enable tardigrade-snail contact?To check whether tardigrades may come into a close encounter with the snail in the natural environment (which would be impossible if the tardigrades were only present in the lower layers of the moss), we investigated the distribution of water bears within moss cushions. The observations were performed for 6 samples of dried moss cushions (ca. 1 cm high and 3 cm in diameter). The moss containing M. inceptum specimens, was collected from a concrete wall in Poznań, Poland (52°24′15″N, 16°53′18″E), the same from which tardigrades were initially collected for the culturing purposes. Three moss cushions were rehydrated, and left for 3 h followed by further observation to check whether tardigrades may actively move across the moss cushion. On the remaining three moss samples, a horizontal cut was made through the center of the moss cushion to check in which layer tardigrade tuns are present while the moss remains dry. The extraction of tardigrades from separated layers was conducted under stereomicroscope (Olympus SZ51) using standard methods33.Within the dry moss cushions tardigrades were present in both the upper and lower moss layers. We did not observe any difference in the number of individuals of M. inceptum that would be dependent on the moss layer. A total of 353 tardigrades were extracted from one moss cushion (dry weight of moss = 0.332 g), what gives the density of tardigrades per 1 g of dry moss sample equal to 1063 specimens. The observation of rehydrated moss cushions conducted in vivo using Olympus SZX16 stereomicroscope associated with Olympus DP74 digital camera and cellSens software revealed that single active tardigrades may also appear on the moss surface (Fig. 1C, red arrow). Therefore, observed in the pilot studies tardigrades distribution within the moss cushion enables tardigrade-snail contact.Is it possible for a tardigrade to take a snail ride?The initial observations were carried out for snails and tardigrades to check whenever a tardigrade may be transferred by a snail. In total, 10 snails and 20 active tardigrades were used. Two variants of Petri dishes (ø 90 mm) were prepared: (1) with smooth and (2) scratched bottom, to avoid and allow tardigrade attachment to the bottom of the dish, respectively. We repeated the observation five times per option. For each single observation we used one snail and two tardigrades.Snails and tardigrades were split equally between the pilot’s experimental options (in total 5 snails and 10 tardigrades per option). We checked whether tardigrades may be transferred by snails by putting tardigrades in the drop of water in the center of a Petri dish and releasing an active snail to crawl through the drop. In total, in the case of the smooth-bottom option, three tardigrades glued to the snail’s body within which two were moved to a distance up to a few centimeters. The third one fixed to a snail’s leg and had a potential to be transferred to a greater distance. In the case of the dishes with the scratched bottom, we did not notice any transfer. Tardigrades were attached tightly to the dishes’ bottom and remained unmoved after the snail had passed through them. Therefore, the observation in the pilot study confirmed that tardigrades may stick to snails’ body and be transferred by a gastropod at least when the substratum (bottom of the dish) is smooth.Experimental design
Experiment 1. Do snails have a significant effect on tardigrade dispersion that depends on the substrate type?As the laboratory environment offers limited possibilities to reflect natural conditions, we aimed to create an environment similar to the natural one by eliminating as many artificial elements as possible and, at the same time, enabling observation and data collection. To imitate a natural microhabitat of water bears we used a piece of moss as a substrate. Moss is a natural shelter and a hunting space for these animals, and a gripping surface that prevents them from being easily carried away by a stream of water or wind. The moss Vesicularia dubyana39 used in the experiment was purchased in an aquarium shop and was derived from an in vitro culture. It was checked under Olympus SZX7 stereomicroscope prior to the experiment to ensure it was free of tardigrades. For experimental purposes we used plastic ventilated boxes with dimensions 950 mm × 950 mm × 600 mm, tightly closed with a plastic lid. The bottom of each box was scratched with sandpaper in order to (1) imitate a rough surface of a concrete wall to which mosses are attached in the natural environment; (2) allow tardigrade locomotion. At the same time, moss and (unfortunately) plastic elements are quite common surroundings of C. nemoralis frequently found in anthropogenic habitats36.Using transparent, non-toxic aquarium silicone, a square with a side length of 3 cm and a height of 0.5 cm was mounted on the bottom of the box. Before starting the experiment, the tightness of the square silicone barrier was checked by pouring 2.5 ml of water inside and leaving the boxes for observation for 24 h. After this time, all silicone squares turned out to be impermeable to water.Boxes for each of the experimental option, namely: (A) control (further in the text referred as C), (B) tardigrades + snail (referred as TS), and (C) tardigrades + snail + moss (referred as TSM, see Fig. 2), were prepared in a following way: 2.5 ml of water was added to the scratched bottom of the box inside the silicone square and 7.5 ml to the area outside of the silicone square to enable survival and active locomotion of tardigrades on both sides of the silicone barrier. Then, 10 active individuals of M. inceptum taken from the culture were transferred to the center of the silicone square. It was repeated for 90 boxes (30 boxes per each C, TS and TSM option). Therefore we used 300 tardigrades per each experimental option which gives 900 tardigrades in total for all experimental options. In case of 30 boxes with TSM option, a piece of moss (ca. 2.5 cm in diameter) was added. It was situated in the center of the silicone square, just after the tardigrades were placed at the boxes in order to isolate tardigrades from the snail during the experiment.Figure 2Graphical representation of three designed experimental options of the experiment 1. (A) 10 tardigrades in the silicone square (control (C)); (B) 10 tardigrades in the silicone square and one snail placed in the box (tardigrades + snail (TS)); (C) 10 tardigrades in the silicone square, one snail placed in the box and additional piece of the moss added as a barrier between tardigrades and snail (tardigrades + snail + moss (TSM)). Figures were assembled in Corel Photo-Paint 2017 (http://www.corel.com).Full size imageFinally, in the boxes targeted for TS and TSM experimental options, one adult and active individual of C. nemoralis snail was placed in each box outside the silicone square. In total, 60 snails were used (30 individuals per experimental option).The boxes were then placed in the rearing room (17 °C, 80% of humidity, photoperiod 12:12) for 72 h. After this time, the number of tardigrades inside and outside the silicone square was counted (both: live and dead) separately for each box, using Olympus SZX7 stereomicroscope.Experiment 2. Effect of the snail’s mucus on tardigrade recovery to active life after anhydrobiosis
Milnesium inceptum anhydrobiosis protocolOnly fully active, adult specimens of medium body length were selected for the experiment. The animals were transferred to ø 3.5 cm vented Petri-dishes with bottom scratched by sandpaper to allow tardigrade locomotion. Five tardigrade individuals were placed to each Petri dish together with 450 µl of water and then dehydrated. In total, 16 Petri dishes with 5 tardigrades on each were prepared. Dehydration process lasted 72 h and was performed in the Q-Cell incubator (40–50% RH, 20 °C, darkness). After that time tardigrade tuns were kept under the abovementioned conditions for 7 days.Impact of the snail’s mucus on tardigrade tunsAfter 7 days of anhydrobiosis, one individual of C. nemoralis was transferred to each dish with tardigrade tuns and was left there for 1 min allowing the snail to actively crawl over the tuns. 30 min after the snail was removed from the dish, tardigrade tuns were observed under the Olympus SZX7 stereomicroscope for any animal movements. Then, all covered and vented dishes were left in the Q-Cell incubator overnight. After 24 h, the dried tuns were rehydrated by adding 3 ml of water to each Petri dish to check whether snail’s mucus affected mortality rates of tardigrades. After 3 and 24 h following rehydration tardigrade tuns were observed for any animal movements. Pictures of tuns were taken using Olympus SZ61 stereomicroscope associated with Olympus UC30 camera (Fig. 3). As reference data on the rehydration of the M. inceptum tuns free of the snail’s mucus, we used the data from Roszkowska et al.20 who tested anhydrobiosis survivability of above-mentioned species. Individuals used for the tuns preparation in the control option were collected from the same laboratory breeding stock, and prepared at the same laboratory conditions as those used in our experiments20.Figure 3Milnesium inceptum tuns: (A,B) before contact with snail mucus; (C,D) coated with wet snail mucus; (E,F) coated with dry snail mucus. Figures were assembled in Corel Photo-Paint 2017 (http://www.corel.com).Full size imageStatistical analysesThe number of tardigrades relocated in each experimental option (C, TS and TSM) was compared with a one-way ANOVA randomized version using RundomPro 3.14 software40. We used non-parametric methods because of the lack of normality. Differences were considered significant at p  More

Boyle, T. H. & Anderson, E. in Cacti: Biology and Uses (ed. Nobel, P. S.) 125–141 (Univ. California Press, 2002).Gibson, A. C. & Nobel, P. S. The Cactus Primer (Harvard Univ. Press, 1986).Bravo Hollis, H. & Sánchez Mejorada, H. Las Cactáceas de México (Univ. Nacional Autónoma de México, 1978).Goettsch, B. et al. High proportion of cactus species threatened with extinction. Nat. Plants 1, 15142 (2015).CAS 
PubMed 

Google Scholar 
Benavides, E., Breceda, A. & Anadón, J. D. Winners and losers in the predicted impact of climate change on cacti species in Baja California. Plant Ecol. 222, 29–44 (2021).
Google Scholar 
Nobel, P. S. Responses of some North American CAM plants to freezing temperatures and doubled CO2 concentrations: implications of global climate change for extending cultivation. J. Arid. Environ. 34, 187–196 (1996).
Google Scholar 
Reyes-García, C. & Andrade, J. L. Crassulacean acid metabolism under global climate change. N. Phytol. 181, 754–757 (2009).
Google Scholar 
Smith, S. D., Didden-Zopfy, B. & Nobel, P. S. High-temperature responses of North American cacti. Ecology 65, 643–651 (1984).
Google Scholar 
Larios, E., González, E. J., Rosen, P. C., Pate, A. & Holm, P. Population projections of an endangered cactus suggest little impact of climate change. Oecologia 192, 439–448 (2020).PubMed 

Google Scholar 
Esparza-Olguı́n, L., Valverde, T. & Vilchis-Anaya, E. Demographic analysis of a rare columnar cactus (Neobuxbaumia macrocephala) in the Tehuacan Valley, Mexico. Biol. Conserv. 103, 349–359 (2002).
Google Scholar 
Seal, C. E. et al. Thermal buffering capacity of the germination phenotype across the environmental envelope of the Cactaceae. Glob. Change Biol. 23, 5309–5317 (2017).
Google Scholar 
Huang, J., Yu, H., Guan, X., Wang, G. & Guo, R. Accelerated dryland expansion under climate change. Nat. Clim. Change 6, 166–171 (2016).
Google Scholar 
Gurvich, D. E. et al. Combined effect of water potential and temperature on seed germination and seedling development of cacti from a mesic Argentine ecosystem. Flora 227, 18–24 (2017).
Google Scholar 
Nuzhyna, N., Baglay, K., Golubenko, A. & Lushchak, O. Anatomically distinct representatives of Cactaceae Juss. family have different response to acute heat shock stress. Flora 242, 137–145 (2018).
Google Scholar 
Andrade, J. L. & Nobel, P. S. Microhabitats and water relations of epiphytic cacti and ferns in a lowland neotropical forest. Biotropica 29, 261–270 (1997).
Google Scholar 
Williams, D. G., Hultine, K. R. & Dettman, D. L. Functional trade-offs in succulent stems predict responses to climate change in columnar cacti. J. Exp. Bot. 65, 3405–3413 (2014).PubMed 

Google Scholar 
Aragón-Gastélum, J. L. et al. Induced climate change impairs photosynthetic performance in Echinocactus platyacanthus, an especially protected Mexican cactus species. Flora Morphol. Distrib. Funct. Ecol. Plants 209, 499–503 (2014).
Google Scholar 
Martorell, C., Montañana, D. M., Ureta, C. & Mandujano, M. C. Assessing the importance of multiple threats to an endangered globose cactus in Mexico: cattle grazing, looting and climate change. Biol. Conserv. 181, 73–81 (2015).
Google Scholar 
Dávila, P., Téllez, O. & Lira, R. Impact of climate change on the distribution of populations of an endemic Mexican columnar cactus in the Tehuacán-Cuicatlán Valley, Mexico. Plant Biosyst. 147, 376–386 (2013).
Google Scholar 
Conver, J. L., Foley, T., Winkler, D. E. & Swann, D. E. Demographic changes over >70 yr in a population of saguaro cacti (Carnegiea gigantea) in the northern Sonoran Desert. J. Arid. Environ. 139, 41–48 (2017).
Google Scholar 
Carrillo-Angeles, I. G., Suzán-Azpiri, H., Mandujano, M. C., Golubov, J. & Martínez-Ávalos, J. G. Niche breadth and the implications of climate change in the conservation of the genus Astrophytum (Cactaceae). J. Arid. Environ. 124, 310–317 (2016).
Google Scholar 
de Cavalcante, A. M. B. & de Duarte, A. S. Modeling the distribution of three cactus species of the Caatinga biome in future climate scenarios. Int. J. Ecol. Environ. Sci. 45, 191–203 (2019).
Google Scholar 
de Cavalcante, A. M. B., de Duarte, A. S. & Ometto, J. P. H. B. Modeling the potential distribution of Epiphyllum phyllanthus (L.) Haw. under future climate scenarios in the Caatinga biome. An. Acad. Bras. Cienc. 92, 351–358 (2020).
Google Scholar 
Tellez-Valdes, O. & DiVila-Aranda, P. Protected areas and climate change: a case study of the cacti in the Tehuacan-Cuicatlan biosphere reserve, Mexico. Conserv. Biol. 17, 846–853 (2003).
Google Scholar 
dos Santos Simões, S., Zappi, D., da Costa, G. M., de Oliveira, G. & Aona, L. Y. S. Spatial niche modelling of five endemic cacti from the Brazilian Caatinga: past, present and future. Austral Ecol. 45, 1–13 (2019).
Google Scholar 
Gorostiague, P., Sajama, J. & Ortega-Baes, P. Will climate change cause spatial mismatch between plants and their pollinators? A test using Andean cactus species. Biol. Conserv. 226, 247–255 (2018).
Google Scholar 
Butler, C. J., Wheeler, E. A. & Stabler, L. B. Distribution of the threatened lace hedgehog cactus (Echinocereus reichenbachii) under various climate change scenarios. J. Torre. Bot. Soc. 139, 46–55 (2012).
Google Scholar 
Johnson, C. N. Species extinction and the relationship between distribution and abundance. Nature 394, 272–274 (1998).CAS 

Google Scholar 
Thuiller, W., Lavorel, S. & Araújo, M. B. Niche properties and geographical extent as predictors of species sensitivity to climate change. Glob. Ecol. Biogeogr. 14, 347–357 (2005).
Google Scholar 
Enquist, B. J. Cyberinfrastructure for an integrated botanical information network to investigate the ecological impacts of global climate change on plant biodiversity. Preprint at PeerJ https://doi.org/10.7287/peerj.preprints.2615v2 (2016).Buisson, L., Thuiller, W., Casajus, N., Lek, S. & Grenouillet, G. Uncertainty in ensemble forecasting of species distribution. Glob. Change Biol. 16, 1145–1157 (2010).
Google Scholar 
Thuiller, W., Guéguen, M., Renaud, J., Karger, D. N. & Zimmermann, N. E. Uncertainty in ensembles of global biodiversity scenarios. Nat. Commun. 10, 1446 (2019).PubMed 
PubMed Central 

Google Scholar 
Goettsch, B., Durán, A. P. & Gaston, K. J. Global gap analysis of cactus species and priority sites for their conservation. Conserv. Biol. 33, 369–376 (2018).PubMed 

Google Scholar 
Maitner, B. S. et al. The bien R package: A tool to access the Botanical Information and Ecology Network (BIEN) database. Methods Ecol. Evol. 9, 373–379 (2018).
Google Scholar 
Karger, D. N. et al. Climatologies at high resolution for the Earth’s land surface areas. Sci. Data 4, 170122 (2017).PubMed 
PubMed Central 

Google Scholar 
Sanderson, B. M., Knutti, R. & Caldwell, P. A representative democracy to reduce interdependency in a multimodel ensemble. J. Clim. 28, 5171–5194 (2015).
Google Scholar 
Brodzik, M. J., Billingsley, B., Haran, T., Raup, B. & Savoie, M. H. EASE-Grid 2.0: Incremental but significant improvements for Earth-gridded data sets. ISPRS Int. J. Geo-Inf. 1, 32–45 (2012).
Google Scholar 
Venter, O. et al. Global terrestrial human footprint maps for 1993 and 2009. Sci. Data 3, 160067 (2016).PubMed 
PubMed Central 

Google Scholar 
Phillips, S. maxnet: Fitting ‘maxent’ species distribution models with ‘glmnet’. R package version 0.1.4. https://CRAN.R-project.org/package=maxnet (2017).Friedman, J., Hastie, T. & Tibshirani, R. Regularization paths for generalized linear models via coordinate descent. J. Stat. Softw. 33, 1–22 (2010).PubMed 
PubMed Central 

Google Scholar 
Dormann, C. F. et al. Collinearity: a review of methods to deal with it and a simulation study evaluating their performance. Ecography 36, 27–46 (2013).
Google Scholar 
Franklin, S. B., Gibson, D. J., Robertson, P. A., Pohlmann, J. T. & Fralish, J. S. Parallel analysis: a method for determining significant principal components. J. Veg. Sci. 6, 99–106 (1995).
Google Scholar 
Roberts, D. R. et al. Cross-validation strategies for data with temporal, spatial, hierarchical, or phylogenetic structure. Ecography 40, 913–929 (2017).
Google Scholar 
Merow, C., Smith, M. J. & Silander, J. A. A practical guide to MaxEnt for modeling species’ distributions: what it does, and why inputs and settings matter. Ecography 36, 1058–1069 (2013).
Google Scholar 
Allouche, O., Tsoar, A. & Kadmon, R. Assessing the accuracy of species distribution models: prevalence, kappa and the true skill statistic (TSS). J. Appl. Ecol. 43, 1223–1232 (2006).
Google Scholar 
Calabrese, J. M., Certain, G., Kraan, C. & Dormann, C. F. Stacking species distribution models and adjusting bias by linking them to macroecological models. Glob. Ecol. Biogeogr. 23, 99–112 (2014).
Google Scholar 
R Core Team R: A Language and Environment for Statistical Computing Version 3.6.0 (R Foundation for Statistical Computing, 2019). https://www.R-project.org/ More

Xu, Y. et al. Annual oil palm plantation maps in Malaysia and Indonesia from 2001 to 2016. Earth Syst. Sci. Data 12, 847–867 (2020).Article 

Google Scholar 
Meijaard, E. et al. The environmental impacts of palm oil in context. Nat. Plants 6, 1418–1426 (2020).Article 

Google Scholar 
Guillaume, T. et al. Carbon costs and benefits of Indonesian rainforest conversion to plantations. Nat. Commun. 9, 2388 (2018).Article 

Google Scholar 
Ordway, E. M. & Asner, G. P. Carbon declines along tropical forest edges correspond to heterogeneous effects on canopy structure and function. Proc. Natl Acad. Sci. USA 117, 7863–7870 (2020).CAS 
Article 

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

Google Scholar 
Santoro, M. et al. The global forest above-ground biomass pool for 2010 estimated from high-resolution satellite observations. Earth Syst. Sci. Data 13, 3927–3950 (2021).Article 

Google Scholar 
The World Database on Protected Areas (WDPA) (UNEP-WCMC and IUCN, accessed 12 February 2020); www.protectedplanet.netMahmud, A., Rehrig, M. & Hills, G. Improving the Livelihoods of Palm Oil Smallholders: The Role of the Private Sector (FSG, 2010).Lasco, R. Forest carbon budgets in Southeast Asia following harvesting and land cover change. Sci. China 45, 55–64 (2002).Article 

Google Scholar 
Historical Greenhouse Gas Emissions (Climate Watch, accessed 6 October 2021); https://www.climatewatchdata.org/Euler, M., Schwarze, S., Siregar, H. & Qaim, M. Oil palm expansion among smallholder farmers in Sumatra, Indonesia. J. Agric. Econ. 67, 658–676 (2016).Article 

Google Scholar 
Donofrio, S., Rothrock, P. & Leonard, J. J. F. T. Supply Change: Tracking Corporate Commitments to Deforestation-free SupplyChains, 2017 (Forest Trends, 2017).Rist, L., Feintrenie, L. & Levang, P. The livelihood impacts of oil palm: smallholders in Indonesia. Biodivers. Conserv. 19, 1009–1024 (2010).Article 

Google Scholar 
Saadun, N. et al. Socio-ecological perspectives of engaging smallholders in environmental-friendly palm oil certification schemes. Land Use Policy 72, 333–340 (2018).Article 

Google Scholar 
Hansen, M. C., Stehman, S. V. & Potapov, P. V. Quantification of global gross forest cover loss. Proc. Natl Acad. Sci. USA 107, 8650 (2010).CAS 
Article 

Google Scholar 
Santoro, M. & Cartus, O. ESA Biomass Climate Change Initiative (Biomass_cci): Global datasets of forest above-ground biomass for the year 2017 v.1 (Centre for Environmental Data Analysis, 2019); https://doi.org/10.5285/bedc59f37c9545c981a839eb552e4084Busch, J. et al. Reductions in emissions from deforestation from Indonesia’s moratorium on new oil palm, timber, and logging concessions. Proc. Natl Acad. Sci. USA 112, 1328–1333 (2015).CAS 
Article 

Google Scholar 
McGarigal, K., Cushman, S. A. & Ene, E. FRAGSTATS v.4: spatial pattern analysis program for categorical and continuous maps (Univ. Massachusetts, 2012). More

Arneth, A. et al. Terrestrial biogeochemical feedbacks in the climate system. Nat. Geosci. 3, 525–532 (2010).CAS 
Article 

Google Scholar 
Green, J. K. et al. Regionally strong feedbacks between the atmosphere and terrestrial biosphere. Nat. Geosci. 10, 410–414 (2017).CAS 
Article 

Google Scholar 
Heimann, M. & Reichstein, M. Terrestrial ecosystem carbon dynamics and climate feedbacks. Nature 451, 289–292 (2008).CAS 
Article 

Google Scholar 
Beer, C. et al. Temporal and among-site variability of inherent water use efficiency at the ecosystem level. Glob. Biogeochem. Cycles 23, 1–13 (2009).Article 

Google Scholar 
Keenan, T. F. et al. Increase in forest water-use efficiency as atmospheric carbon dioxide concentrations rise. Nature 499, 324–327 (2013).CAS 
Article 

Google Scholar 
Frank, D. C. et al. Water-use efficiency & transpiration across European forests during the Anthropocene. Nat. Clim. Change 5, 579–583 (2015).CAS 
Article 

Google Scholar 
Mastrotheodoros, T. et al. Linking plant functional trait plasticity and the large increase in forest water use efficiency. J. Geophys. Res. Biogeosci. 122, 2393–2408 (2017).Article 

Google Scholar 
Lavergne, A. et al. Observed and modelled historical trends in the water-use efficiency of plants and ecosystems. Glob. Change Biol. 25, 2242–2257 (2019).Article 

Google Scholar 
Huxman, T. E. et al. Convergence across biomes to a common rain-use efficiency. Nature 429, 651–654 (2004).CAS 
Article 

Google Scholar 
Yang, Y. et al. Contrasting responses of water use efficiency to drought across global terrestrial ecosystems. Sci. Rep. 6, 23284 (2016).CAS 
Article 

Google Scholar 
Huang, L. et al. A global examination of the response of ecosystem water-use efficiency to drought based on MODIS data. Sci. Total Environ. 601–602, 1097–1107 (2017).Article 

Google Scholar 
Reichstein, M. et al. Severe drought effects on ecosystem CO2 and H2O fluxes at three Mediterranean evergreen sites: revision of current hypotheses? Glob. Change Biol. 8, 999–1017 (2002).Article 

Google Scholar 
Reichstein, M. et al. Inverse modeling of seasonal drought effects on canopy CO2/H2O exchange in three Mediterranean ecosystems. J. Geophys. Res. Atmos. 108, 4726 (2003).Article 

Google Scholar 
Cooley, S. S. et al. Assessing regional drought impacts on vegetation and evapotranspiration: a case study in Guanacaste, Costa Rica. Ecol. Appl. 29, e01834 (2019).Article 

Google Scholar 
Medrano, H., Flexas, J. & Galmés, J. Variability in water use efficiency at the leaf level among Mediterranean plants with different growth forms. Plant Soil 317, 17–29 (2008).Article 

Google Scholar 
Soh, W. K. et al. Rising CO2 drives divergence in water use efficiency of evergreen and deciduous plants. Sci. Adv. 5, eaax7906 (2019).CAS 
Article 

Google Scholar 
Wang, M., Chen, Y., Wu, X. & Bai, Y. Forest-type-dependent water use efficiency trends across the northern hemisphere. Geophys. Res. Lett. 45, 8283–8293 (2018).Article 

Google Scholar 
Enquist, B. et al. Scaling from traits to ecosystems: developing a general trait driver theory via integrating trait-based and metabolic scaling theories. Adv. Ecol. Res. 52, 249–318 (2015).Article 

Google Scholar 
Gross, N. et al. Functional trait diversity maximizes ecosystem multifunctionality. Nat. Ecol. Evol. 1, 0132 (2017).Article 

Google Scholar 
Bagousse‐Pinguet, Y. L. et al. Testing the environmental filtering concept in global drylands. J. Ecol. 105, 1058–1069 (2017).Article 

Google Scholar 
Ponce Campos, G. E. et al. Ecosystem resilience despite large-scale altered hydroclimatic conditions. Nature 494, 349–352 (2013).CAS 
Article 

Google Scholar 
Fisher, J. B. et al. The future of evapotranspiration: global requirements for ecosystem functioning, carbon and climate feedbacks, agricultural management, and water resources. Water Resour. Res. 53, 2618–2626 (2017).Article 

Google Scholar 
Xue, B.-L. et al. Global patterns, trends, and drivers of water use efficiency from 2000 to 2013. Ecosphere 6, art174 (2015).Article 

Google Scholar 
Fisher, J. B. et al. ECOSTRESS: NASA’s next generation mission to measure evapotranspiration from the International Space Station. Water Resour. Res. 56, e2019WR026058 (2020).Article 

Google Scholar 
Higgins, M. A. et al. Geological control of floristic composition in Amazonian forests. J. Biogeogr. 38, 2136–2149 (2011).Article 

Google Scholar 
De Kauwe, M. G., Keenan, T. F., Medlyn, B. E., Prentice, I. C. & Terrer, C. Satellite based estimates underestimate the effect of CO2 fertilization on net primary productivity. Nat. Clim. Change 6, 892–893 (2016).Article 

Google Scholar 
Huang, M. et al. Seasonal responses of terrestrial ecosystem water-use efficiency to climate change. Glob. Change Biol. 22, 2165–2177 (2016).Article 

Google Scholar 
Lin, Y.-S. et al. Optimal stomatal behaviour around the world. Nat. Clim. Change 5, 459–464 (2015).CAS 
Article 

Google Scholar 
Medlyn, B. E. et al. Reconciling the optimal and empirical approaches to modelling stomatal conductance. Glob. Change Biol. 17, 2134–2144 (2011).Article 

Google Scholar 
Peters, W. et al. Increased water-use efficiency and reduced CO2 uptake by plants during droughts at a continental scale. Nat. Geosci. 11, 744–748 (2018).CAS 
Article 

Google Scholar 
Cheng, L. et al. Recent increases in terrestrial carbon uptake at little cost to the water cycle. Nat. Commun. 8, 110 (2017).Article 

Google Scholar 
Fisher, J. B. ECOsystem Spaceborne Thermal Radiometer Experiment on Space Station (ECOSTRESS): Level-3 Evapotranspiration L3(ET_PT-JPL) Algorithm Theoretical Basis Document. Jet Propulsion Laboratory, California Institute of Technology (2018).Running, S. W. et al. A continuous satellite-derived measure of global terrestrial primary production. BioScience 54, 547–560 (2004).Article 

Google Scholar 
Heinsch, F. et al. Evaluation of remote sensing based terrestrial productivity from MODIS using regional tower eddy flux network observations. IEEE Trans. Geosci. Remote Sens. 44, 1908–1925 (2006).Article 

Google Scholar 
Zhao, M., Heinsch, F., Nemani, R. & Running, S. Improvements of the MODIS terrestrial gross and net primary production global data set. Remote Sens. Environ. 95, 164–176 (2005).Article 

Google Scholar 
Ryu, Y. et al. Integration of MODIS land and atmosphere products with a coupled-process model to estimate gross primary productivity and evapotranspiration from 1 km to global scales. Glob. Biogeochem. Cycles 25, GB4017 (2011).Article 

Google Scholar  More

Systematic palaeontology
Tetrapoda Jaekel, 1909 fide Sues20
Family undesignated
Termonerpeton makrydactylus gen. et sp. nov. (Fig. 1)Fig. 1: Termonerpeton makrydactylus gen. et sp. nov. holotype UMZC 2019.1.a Specimen photograph. b Interpretive drawing. Scale bars 10 mm. Abbreviations: acet acetabulum, fem femur, fib fibula, ha haemal arch, ic intercentrum, l left, na neural arch, phal phalanx, piliac p post-iliac process, plc pleurocentrum, r right, sac rib sacral rib, tib tibia.Full size image

EtymologyGenus: from τέρμωυ (térmon) meaning boundary and ερπετό (erpetó) meaning ‘crawler’, referring to the field boundary walls near the East Kirkton quarry where the late Stan Wood initially discovered fossils from the East Kirkton Limestone and from where the type specimen may have been collected; species: from μακρύς (makrýs) meaning ‘elongate’ and δάχτυλο (dáchtylo; more precisely, δάχτυλο ποδιού, dáchtylo podioú) meaning ‘toe’, referring to the very long pedal digit IV.HolotypeUniversity of Cambridge Museum of Zoology (UMZC) 2019.1. A partial tetrapod postcranium, preserving both pelves, a femur, fibula, tibia, and an almost complete but disarticulated pes. Closely associated with the appendicular elements are dorsally open hoop-shaped centra, a few neural and haemal arches, curved ribs, and a section of articulated gastralia.Locality and horizonEast Kirkton quarry, near Bathgate, Scotland, UK. East Kirkton Limestone, Bathgate Hills Volcanic Formation. Exact horizon is unknown. Brigantian, Viséan, early Carboniferous (=Mississippian)21.Differential diagnosisPossible autapomorphies: ilium with drawn-out, flat, blade-like dorsal process; very large, stout, and elongate metatarsal IV, greatly exceeding the length of metatarsals III and V (~30% or more). Possible tetrapod synapomorphies among post-Devonian taxa: distinct interepipodial space between tibia and fibula; well-ossified tarsus comprising tibiale, fibulare, intermedium, four centralia, and five distal tarsals. Possible amniote synapomorphies, but often showing reversed polarity in several stem- and crown amniote taxa: presumed pedal phalangeal formula 23454; robust and long pedal digit IV; enlarged intermedium and fibulare, together occupying more than half of proximal moiety of tarsus; curved ribs. Characters of uncertain polarity (also present in Caerorhachis): elongate, slender, and posterodorsally oblique post-iliac process; short puboischiadic plate with almost vertical anterior margin; stout femur with poorly pronounced waisting along the shaft, longer than puboischiadic plate; hoop-shaped centra.Attributed specimenNational Museums Scotland (NMS) G.1992.22.1. An articulated, partially complete, large tetrapod pes, preserving a nearly complete array of tarsals, all metatarsals, and the proximal phalanges of digits I–III. Unit 82, East Kirkton Limestone, East Kirkton quarry, near Bathgate, Scotland, UK.
Specimen description
Appendicular skeletonMost of the description is based upon the holotype. Both pelves are preserved, one mainly as a natural mould. The puboichiadic plates are short and deep, with an almost vertical anterior margin to the pubis (Fig. 1). In one, the surface of the puboischiadic plate is strongly convex, in the other it is strongly concave. The concave plate may belong to the left pelvis, with the concavity indicating the acetabulum. Both iliac processes of the presumed right ilium are overlain by a neural arch and part of the femur and cannot be seen. The presumed left ilium shows a long, posteriorly pointing post-iliac process that extends as far backward as the posterior edge of the ischium. It retains the proximal, stump-like portion of a dorsal iliac process, continued distally in natural mould as a mediolaterally flattened and blade-like structure. Both processes sit above a short iliac neck. The dorsal iliac process is proportionally longer than in other tetrapods and its knife-like appearance is unique. The angle between the two processes is much more acute than in most other tetrapods, and the nearest comparison is with the divided iliac process of the microsaurs Tuditanus and Ricnodon22 which, however, could merely represent a bifid post-iliac process. Two gaps in ossification are taken as evidence of an ilio-ischiadic suture half-way down the posterior margin on the left pelvis and an ilio-pubic suture halfway down the anterior margin of the right pelvis (Fig. 1). There is no evidence of a puboischiadic suture, although a shallow depression along the ventral margin of the left puboischiadic plate probably marks the junction between pubis and ischium. The complete left puboischiadic plate is 20 mm deep behind the ilium and 30 mm long, with the pubis contributing about one-third of its length and the ischium the remaining two thirds. The anterior margin of the pubis is almost vertical. The dorsal margin of the ischium is shallowly convex for half its length before extending posteroventrally to meet the upturned posterior extremity of the ischium’s ventral margin. There is no evidence as to the angle at which the two pelvic plates met at the symphysis, which would affect the position of the acetabulum relative to the substrate, and thus the effective resting posture of the hindlimb.The left femur is at least 39 mm in length, and longer than the puboischiadic plate. The entire bone is crushed, and its distal end lies partly beneath one of the pelvic halves and a neural arch so that its features cannot easily be made out. A possible intercondylar groove may be present distally, and the extensor surface of its proximal extremity appears to show a subcentral depression. The femur itself is robust with little waisting at mid-shaft. A small internal trochanter lies near its proximal end. The left fibula is approximately 26 mm long along its lateral margin. Its proximal end is narrow and grooved. Its broad and strongly flared distal end suggests a broad contact with the tarsus. The medial turn of the distal end indicates a large interepipodial space. The left tibia is about 20 mm long, slender, and shallowly waisted at mid-shaft. It is not clear which end is proximal and which distal, although probably the proximal is the broader. The tibia is probably more than half the length of the femur. Based upon the femur and tibia lengths, and omitting the ankle and pes, the above figures indicate a total stylopod-zeugopod length of about 65 mm, assuming a fully extended limb.Most of the morphology of the left pes is preserved, showing many well-ossified tarsal bones (Fig. 2). Several of these, including possible distal tarsals II and III lie more or less in anatomical continuity relative to metatarsals II and III, respectively. Other tarsal elements, including possible fibulare, tibiale, centralia, and distal tarsals, are illustrated in Fig. 2. Metatarsal IV lies in anatomical position relative to metatarsals II and III and, at 7 mm in length, it is significantly larger than the latter. The presumed first phalanx of pedal digit IV lies close to metatarsal IV, at an angle of nearly 90° to the latter. It is long and slender, indicating an unusually elongate fourth pedal digit.Fig. 2: Termonerpeton makrydactylus gen. et sp. nov. left hindlimb of UMZC 2019.1.a Specimen photograph, showing close-up view of hindlimb skeleton, b Interpretive drawing, with centralia, distal tarsals, and metatarsals indicated by red, blue, and black Roman numerals, respectively, c Interpretive drawing with dashed lines connecting elements of individual digits, d Reconstruction of left tibia, fibula and pes. Scale bars 10 mm. Abbreviations: interm intermedium, tib tibialia.Full size imageAn array of about 12 phalanges is preserved. They are all disrupted but occur in proximity to one another and, like the first phalanx of pedal digit IV, also mainly lie at right angles to metatarsals III and IV. An additional, acutely angled pointed ungual phalanx, possibly associated with digit II, is also visible. A further two phalanges have been displaced and rest along the anterior edge of the left pelvis. In total, we were, therefore, able to identify 15 elements. The preservation of the pes suggests it was strongly flexed either at death or from tissue shrinkage thereafter. An isolated metatarsal, presumably from the other, missing foot, lies some distance away near the edge of the block. Together, the pedal elements suggest a relatively large foot.A second specimen, NMS G.1992.22.1 (Fig. 3), is represented by an isolated pes. It may belong to Termonerpeton, although it is from a much larger individual. It shows five metatarsals of which the fourth is much longer and more robust than the other four and about twice as long as that of the holotype, while metatarsal V is the smallest. There are three phalanges, plus five distal tarsals. A D-shaped element closely associated with three centralia could be either a fibulare, a displaced intermedium, or centrale IV.Fig. 3: Termonerpeton makrydactylus gen. et sp. nov partial pes, attributed specimen NMS G.1992.22.1.a Specimen photograph, b with centralia, distal tarsals, and metatarsals indicated by red, blue, and black Roman numerals, respectively. Scale bars 10 mm.Full size imageAxial skeletonWhere visible, neural arches have short neural spines and prominent zygapophyses, but their shape is hard to assess as none is well preserved. The element overlying part of the right pelvis and the femur is 7 mm high in total. Numerous dorsally open, hoop-shaped centra about 5 mm in diameter are visible, as well as a few small, oval, shallowly curved elements (Fig. 1). Without further evidence, it is uncertain which of these elements are intercentra and which pleurocentra, though we assume that the larger elements are pleurocentra. The preserved ribs are slender and curved, and include trunk ribs, a possible presacral rib, a possible sacral rib, and a possible postsacral rib. This is long but more or less straight. A bone situated among a cluster of centra, somewhat distant from the other tarsal bones, was originally interpreted by us as a possible fibulare, similar to the fibulare in Proterogyrinus23. However, it might also be interpreted as a sacral rib. If so, its morphology is unique. It is short and widens distally into a fan-shaped structure but does not appear to have a bifid proximal end, unlike the sacral rib in Proterogyrinus23. Three haemal arches are present, one still attached to its half-hoop centrum, a second slightly longer, and a third very short and presumably from a more posterior region of the tail.ComparisonsThe exceptional preservation of tetrapods from the East Kirkton Limestone provides a unique opportunity to study portions of the skeletal anatomy that are otherwise poorly preserved or absent among Mississippian tetrapods. In particular, hindlimbs with a complete or near-complete array of tarsal elements and digits are notably rare. The unusual construction of the pes of Termonerpeton prompted us to examine the hindlimb morphology of six other East Kirkton tetrapods (Fig. 4a–g) alongside a selection of additional, mostly Carboniferous taxa (Fig. 4h–n). We focus on epipodials, tarsi, phalangeal formulae and digit length and proportions. To facilitate visual inspection of these elements, all hindlimbs are drawn to a common tibial length, except for the stem diapsid Petrolacosaurus, in which the epipodials are greatly elongate.Fig. 4: Comparison of the left tibia, fibula, tarsus, and digits of early tetrapods.a Balanerpeton after 2, b Eucritta after 12, c Eldeceeon after 6, d Silvanerpeton after 4, e Westlothiana after 7, f Kirktonecta original, see 15 (the grey area marks the estimated position and extent of the tarsus), g Termonerpeton, h Pederpes after 24, i, Greererpeton after 27, j Caerorhachis after 31, k Archeria after 30, l Hylonomus after 28, m Tuditanus after 22, n Petrolacosaurus after 29. Drawn to the same tibial length apart from n. Scale bars 10 mm.Full size imageIn terms of pes size relative to the tibia, the East Kirkton taxa Balanerpeton, Eucritta, and Silvanerpeton (Fig. 4a, b, d) are similarly proportioned. In contrast, Eldeceeon and Westlothiana (Fig. 4c, e) exhibit somewhat larger pedes. Kirktonecta has proportionally the largest pedes of all (Fig. 4f). Termonerpeton (Fig. 4g) has a pes of similar size to the first three taxa except that digit IV is relatively much longer than in any of the others, with an exceptionally large metatarsal IV. In all those taxa in which digit IV is fully preserved, it is the longest, especially in Eldeceeon and Kirktonecta, but in none does it approach in size and proportions that of Termonerpeton. The illustrated limbs also differ from one another in the degree of ossification of the tarsal bones. Most taxa except Eucritta have some indication of ossified tarsal elements, and some of them, including Balanerpeton and Silvanerpeton, show a complete or almost complete tarsal set. Kirktonecta does have an ossified tarsus, but specimen preservation does not allow us to identify individual elements. The phalangeal count, where known, also varies: 22343 in Balanerpeton2; 223?? in Eucritta12; 23455 in Silvanerpeton4; 23454 in Eldeceeon6, Kirktonecta15, Termonerpeton, and Westlothiana7.In addition, we compared the pedes of East Kirkton tetrapods with those of seven other taxa (Fig. 4h–n): one earlier, Pederpes24; one almost contemporary, Caerorhachis25; four later Carboniferous, Greererpeton26, Hylonomus27, Tuditanus22, and Petrolacosaurus28; and one early Permian, Archeria29. Of these, Greererpeton has relatively the smallest pes. In most, digit IV is the longest, though in Pederpes and Caerorhachis it is incomplete. The pes of Caerorhachis was originally restored with only three phalanges in digit IV30. This is probably incorrect and would be unusual in Carboniferous tetrapods. The pes of the anthracosaur Archeria was originally reconstructed with digit V as the longest29, but again this is unusual among later Carboniferous and early Permian tetrapods and we suspect that digits IV and V have been transposed, and Romer himself expressed doubt about this reconstruction29. In either case, the phalangeal formula of Archeria is similar to that of the East Kirkton anthracosaur Silvanerpeton, as 23455.Among Carboniferous tetrapods, temnospondyls such as Balanerpeton and colosteids such as Greererpeton show a digit IV that is somewhat longer than the others, but metatarsal IV is very similar in length and breadth to the adjacent metatarsals. In anthracosaurs, digit IV is the longest, but again metatarsal IV is not significantly broader than adjacent metatarsals. This is also the case in the early amniote Hylonomus and the microsaur Tuditanus. Among the taxa illustrated here, Termonerpeton shows a strikingly similar pes to that of the Late Pennsylvanian araeoscelidian diapsid Petrolacosaurus (Fig. 4n). In both, metatarsal IV is significantly longer and stouter than others and forms part of a similarly long digit IV. In early amniotes, an elongate digit IV coupled with an elongate metatarsal IV is a common occurrence in other taxa, such as protothyridids (e.g. Anthracodromeus31), basal araeoscelidians (e.g. Spinoaequalis32), younginids (e.g. Youngina33), saurians33, and basal synapsids (e.g. Heleosaurus34,35,36), among others.Based upon available evidence, an elongate digit IV is likely to be the plesiomorphic condition for crown amniotes, being present in Hylonomus, Paleothyris, and Petrolacosaurus (Fig. 4l, n), and shortening of this digit certainly represents a derived feature. In later crown amniotes, the conditions vary, with larger, heavier-bodied tetrapods such as dicynodonts and diadectids having generally shorter toes and adopting a more clearly plantigrade posture. An elongate metatarsal IV and associated digit, however, are not universal among Palaeozoic amniotes, and modifications of these conditions occur repeatedly across clades. For instance, in the eureptile captorhinid Eocaptorhinus, digit IV is also the longest, but the length of metatarsal IV does not greatly exceed that of other metatarsals37. The same is true of some early Permian clades, including seymouriamorphs (e.g. Seymouria38; Discosauriscus39), and diadectids (e.g. Diadectes40), although in the diadectomorph Orobates digit III is a little longer than digit IV41. Among synapsids, dicynodonts such as Diictodon42 and caseids43, to name a few, have five pedal digits of approximately uniform length.We further point out that, while digit IV attains a certain degree of elongation in other early tetrapod groups, such as temnospondyls, in none of them do the relative proportions of this digit (where known) compare to those of several stem and crown amniotes (Fig. 4).Phylogenetic relationshipsThe results of various phylogenetic analyses lend some support to the interpretation of Termonerpeton as a stem amniote, despite its uncertain placement in the unweighted character parsimony analysis (Fig. 5a). In the latter analysis, Termonerpeton appears in a polytomous node alongside baphetids (Eucritta; Baphetes; Megalocephalus), temnospondyls (Balanerpeton; Dendrysekos), the anthracosauroids Eldeceeon and Silvanerpeton, and the problematic Caerorhachis. In all other analyses—implied weights, reweighted characters, and Bayesian—Termonerpeton is placed on the amniote stem group, albeit in different positions, among a diverse array of ‘reptiliomorph’ clades and grades. In the implied weights analysis (Fig. 5b), Termonerpeton, Silvanerpeton, and Eldeceeon form a monophyletic group branching crownward of chroniosaurs plus anthracosaurs and anti-crownward of paraphyletic gephyrostegids. In the reweighted analysis (Fig. 5c), Termonerpeton and Caerorhachis appear as successive sister taxa, in that order, to monophyletic anthracosaurs. In the Bayesian analysis (Fig. 5d), the amniote total group receives moderate support with a credibility value (c.v.) of 76 with Caerorhachis as the most plesiomorphic stem amniote. Crownward of Caerorhachis is a polytomy with low support (c.v. = 59) that subtends Termonerpeton, a clade consisting of Eldeceeon plus Silvanerpeton, a clade of anthracosaurs, and a clade that includes all remaining taxa. In crownward succession, these taxa include chroniosaurs, gephyrostegids, seymouriamorphs, Solenodonsaurus, and Westlothiana as successive sister groups to a strongly supported (c.v. = 100) clade containing diadectomorphs, synapsids, and eureptiles. Although eureptile monophyly is not retrieved, strong support (c.v. = 100) is given to the branch subtending diadectomorphs plus synapsids44.Fig. 5: Results of phylogenetic analyses.a Strict consensus of 120 shortest trees from unweighted analysis (tree length = 1286 steps, ensemble consistency index C.I. = 0.2738 without uninformative characters, ensemble retention index R.I. = 0.5768), b Single tree from implied weights analysis (tree length = 1298 steps, Goloboff fit = −202.59266, C.I. = 0.2712, R.I. = 0.5713), c Single tree from reweighted analysis (tree length = 212,68965 steps, C.I. = 0.4755, R.I. = 0.774), d Bayesian topology with branches reporting credibility values.Full size image More

Oppel, S. et al. Spatial scales of marine conservation management for breeding seabirds. Mar. Policy 98, 37–46 (2018).Article 

Google Scholar 
Lewison, R. et al. Research priorities for seabirds: improving conservation and management in the 21st century. Endanger. Species Res 17, 93–121 (2012).Article 

Google Scholar 
Hasegawa, H. & DeGange, A. R. The Short-tailed Albatross, Diomedea albatrus, its status, distribution and natural history. Am. Birds 36, 806–814 (1982).
Google Scholar 
Tickell, W. L. N. Albatrosses (Pica Press, 2000).BirdLife International. Phoebastria albatrus. The IUCN Red List of Threatened Species, e.T22698335A132642113 https://doi.org/10.2305/IUCN.UK.2018-2.RLTS.T22698335A132642113.en (2018).Japan Ministry of the Environment. Ministry of the Environment Red List (Government of Japan, 2020).COSEWIC. COSEWIC Assessment and Status Report on the Short-tailed Albatross Phoebastria albatrus in Canada (Committee on the Status of Endangered Wildlife in Canada, 2013).Environment Canada. Recovery Strategy for the Short-tailed Albatross (Phoebastria albatrus) and the Pink-footed Shearwater (Puffinus creatopus) in Canada (Environment Canada, 2008).United States of America Fish and Wildlife Service. Endangered and Threatened Wildlife and Plants; Final Rule to List the Short-tailed Albatross as Endangered in the United States. 65 FR 46643, 46643–4654, Document Number 00–19123 (2000).United States of America Fish and Wildlife Service. Short-tailed Albatross (Phoebastria albatrus) 5-Year Review: Summary and Evaluation (United States of America Fish and Wildlife Service, 2020).United States of America Fish and Wildlife Service. Short-tailed Albaross Recovery Plan (United States of America Fish and Wildlife Service, 2008).Orben, R. A. et al. Ontogenetic changes in at-sea distributions of immature short-tailed albatrosses Phoebastria albatrus. Endanger. Species Res 35, 23–37 (2018).Article 

Google Scholar 
Orben, R. A. et al. Across borders: external factors and prior behaviour influence North Pacific albatross associations with fishing vessels. J. Appl. Ecol. 58, 1272–1283 (2021).Article 

Google Scholar 
Fox, C. H., Robertson, C., O’Hara, P. D., Tadey, R. & Morgan, K. H. Spatial assessment of albatrosses, commercial fisheries, and bycatch incidents on Canada’s Pacific coast. Mar. Ecol. Prog. Ser. 672, 205–222 (2021).Article 

Google Scholar 
Piatt, J. F. et al. Predictable hotspots and foraging habitat of the endangered short-tailed albatross (Phoebastria albatrus) in the North Pacific: implications for conservation. Deep Sea Res. Part II 53, 387–398 (2006).Article 

Google Scholar 
Suryan, R. M. et al. Migratory routes of short-tailed albatrosses: use of exclusive economic zones of North Pacific Rim countries and spatial overlap with commercial fisheries in Alaska. Biol. Conserv. 137, 450–460 (2007).Article 

Google Scholar 
Suryan, R. M. & Fischer, K. N. Stable isotope analysis and satellite tracking reveal interspecific resource partitioning of nonbreeding albatrosses off Alaska. Can. J. Zool. 88, 299–305 (2010).CAS 
Article 

Google Scholar 
Zador, S. G., Punt, A. E. & Parrish, J. K. Population impacts of endangered short-tailed albatross bycatch in the Alaskan trawl fishery. Biol. Conserv. 141, 872–882 (2008).Article 

Google Scholar 
Geernaert, T. O., Gilroy, H. L., Kaimmer, S. M., Williams, G. H. & Trumble, R. J. A Feasibility Study that Investigates Options for Monitoring Bycatch of the Short-tailed Albatross in the Pacific Halibut Fishery off Alaska (International Pacific Halibut Commission, 2001).Guy, T. J. et al. Overlap of North Pacific albatrosses with the U.S. west coast groundfish and shrimp fisheries. Fish. Res. 147, 222–234 (2013).Article 

Google Scholar 
Bolnick, D. I. et al. The ecology of individuals: incidence and implications of individual specialization. Am. Natural 161, 1–28 (2003).Article 

Google Scholar 
Votier, S. C. et al. Individual responses of seabirds to commercial fisheries revealed using GPS tracking, stable isotopes and vessel monitoring systems. J. Appl. Ecol. 47, 487–497 (2010).Article 

Google Scholar 
Wakefield, E. D. et al. Long-term individual foraging site fidelity—why some gannets don’t change their spots. Ecology 96, 3058–3074 (2015).PubMed 
Article 
PubMed Central 

Google Scholar 
Votier, S. C. et al. Effects of age and reproductive status on individual foraging site fidelity in a long-lived marine predator. Proc. R. Soc. B 284, 20171068 (2017).PubMed 
PubMed Central 
Article 

Google Scholar 
Sztukowski, L. A. et al. Sex differences in individual foraging site fidelity of Campbell albatross. Mar. Ecol. Prog. Ser. 601, 227–238 (2018).Article 

Google Scholar 
Gutowsky, S. E. et al. Divergent post-breeding distribution and habitat associations of fledgling and adult Black-footed Albatrosses Phoebastria nigripes in the North Pacific. Ibis 156, 60–72 (2014).Article 

Google Scholar 
Weimerskirch, H., Åkesson, S. & Pinaud, D. Postnatal dispersal of wandering albatrosses Diomedea exulans: implications for the conservation of the species. J. Avian Biol. 37, 23–28 (2006).
Google Scholar 
Olson, S. L. & Hearty, P. J. Probable extirpation of a breeding colony of Short-tailed Albatross (Phoebastria albatrus) on Bermuda by Pleistocene sea-level rise. Proc. Natl Acad. Sci. 100, 12825–12829 (2003).CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Dall, W. H. Notes on pre-historic remains in the Aleutian islands. Proc. Calif. Acad. Sci. 4, 283–287 (1872).
Google Scholar 
Eda, M. et al. Inferring the ancient population structure of the vulnerable albatross Phoebastria albatrus, combining ancient DNA, stable isotope, and morphometric analyses of archaeological samples. Conserv. Genet. 13, 143–151 (2012).Article 

Google Scholar 
Cousins, K. L., Dalzell, P. & Gilman, E. Managing pelagic longline-albatross interactions in the North Pacific Ocean. Mar. Ornithol 28, 159–174 (2000).
Google Scholar 
Hobson, K. A. & Montevecchi, W. A. Stable isotopic determinations of trophic relationships of great auks. Oecologia 87, 528–531 (1991).PubMed 
Article 
PubMed Central 

Google Scholar 
Fuller, B. T. et al. Pleistocene paleoecology and feeding behavior of terrestrial vertebrates recorded in a pre-LGM asphaltic deposit at Rancho La Brea, California. Palaeogeogr. Palaeoclimatol. Palaeoecol. 537, 109383 (2020).Article 

Google Scholar 
Hobson, K. A. & Clark, R. G. Assessing avian diets using stable isotopes I: turnover of 13C in tissues. Condor 94, 181–188 (1992).Article 

Google Scholar 
Hyland, C., Scott, M. B., Routledge, J. & Szpak, P. Stable carbon and nitrogen isotope variability of bone collagen to determine the number of isotopically distinct specimens. J. Archaeol. Method Theory https://doi.org/10.1007/s10816-021-09533-7 (2021).Article 

Google Scholar 
Hedges, R. E. M., Clement, J. G., Thomas, D. L. & O’Connell, T. C. Collagen turnover in the adult femoral mid‐shaft: modeled from anthropogenic radiocarbon tracer measurements. Am. J. Phys. Anthropol. 133, 808–816 (2007).PubMed 
Article 
PubMed Central 

Google Scholar 
Guiry, E. J., Orchard, T. J., Royle, T. C. A., Cheung, C. & Yang, D. Y. Dietary plasticity and the extinction of the passenger pigeon (Ectopistes migratorius). Quat. Sci. Rev. 233, 106225 (2020).Article 

Google Scholar 
Minagawa, M. & Wada, E. Stepwise enrichment of 15N along food chains: further evidence and the relation between δ15N and animal age. Geochim. Cosmochim. Acta 48, 1135–1140 (1984).CAS 
Article 

Google Scholar 
DeNiro, M. J. & Epstein, S. Influence of diet on the distribution of carbon isotopes in animals. Geochim. Cosmochim. Acta 42, 495–506 (1978).CAS 
Article 

Google Scholar 
Hobson, K. A., Ambrose, W. G. Jr & Renaud, P. E. Sources of primary production, benthic-pelagic coupling, and trophic relationships within the Northeast Water Polynya: insights from δ13C and δ15N analysis. Mar. Ecol. Prog. Ser. 128, 1–10 (1995).Article 

Google Scholar 
Sigman, D., Karsh, K. & Casciotti, K. Ocean process tracers: nitrogen isotopes in the ocean in Encyclopedia of Ocean Science (eds Steele, J. H. et al.) 4139–4152 (Academic Press, 2009).Guiry, E. Complexities of stable carbon and nitrogen isotope biogeochemistry in ancient freshwater ecosystems: implications for the study of past subsistence and environmental change. Front. Ecol. Evol 7, 313 (2019).Article 

Google Scholar 
Rau, G. H., Takahashi, T. & Des Marais, D. J. Latitudinal variations in plankton δ13C: implications for CO2 and productivity in past oceans. Nature 341, 516–518 (1989).CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
Popp, B. N. et al. Effect of phytoplankton cell geometry on carbon isotopic fractionation. Geochim. Cosmochim. Acta 62, 69–77 (1998).CAS 
Article 

Google Scholar 
Laws, E. A., Popp, B. N., Bidigare, R. R., Kennicutt, M. C. & Macko, S. A. Dependence of phytoplankton carbon isotopic composition on growth rate and (CO2) aq: theoretical considerations and experimental results. Geochim. Cosmochim. Acta 59, 1131–1138 (1995).CAS 
Article 

Google Scholar 
Vokhshoori, N. L. et al. Broader foraging range of ancient short-tailed albatross populations into California coastal waters based on bulk tissue and amino acid isotope analysis. Mar. Ecol. Prog. Ser. 610, 1–13 (2019).CAS 
Article 

Google Scholar 
Sherwood, O. A., Lehmann, M. F., Schubert, C. J., Scott, D. B. & McCarthy, M. D. Nutrient regime shift in the western North Atlantic indicated by compound-specific δ15N of deep-sea gorgonian corals. Proc. Natl Acad. Sci. 108, 1011–1015 (2011).CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Szpak, P., Savelle, J. M., Conolly, J. & Richards, M. P. Variation in late holocene marine environments in the Canadian Arctic Archipelago: evidence from ringed seal bone collagen stable isotope compositions. Quat. Sci. Rev. 211, 136–155 (2019).Article 

Google Scholar 
Guiry, E. J. et al. Deforestation caused abrupt shift in Great Lakes nitrogen cycle. Limnol. Oceanogr. 65, 1921–1935 (2020).CAS 
Article 

Google Scholar 
Wiley, A. E. et al. Millennial-scale isotope records from a wide-ranging predator show evidence of recent human impact to oceanic food webs. Proc. Natl Acad. Sci. 110, 8972–8977 (2013).CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Keeling, C. D. The Suess effect: 13Carbon-14Carbon interrelations. Environ. Int. 2, 229–300 (1979).CAS 
Article 

Google Scholar 
McMahon, K. W., Thorrold, S. R., Elsdon, T. S. & McCarthy, M. D. Trophic discrimination of nitrogen stable isotopes in amino acids varies with diet quality in a marine fish. Limnol. Oceanogr. 60, 1076–1087 (2015).CAS 
Article 

Google Scholar 
Chikaraishi, Y. et al. Determination of aquatic food‐web structure based on compound‐specific nitrogen isotopic composition of amino acids. Limnol. Oceanogr. Methods 7, 740–750 (2009).CAS 
Article 

Google Scholar 
Jackson, A. L., Inger, R., Parnell, A. C. & Bearhop, S. Comparing isotopic niche widths among and within communities: SIBER–stable isotope Bayesian ellipses in R. J. Anim. Ecol. 80, 595–602 (2011).PubMed 
Article 
PubMed Central 

Google Scholar 
Ambrose, S. H. Preparation and characterization of bone and tooth collagen for isotopic analysis. J. Archaeol. Sci. 17, 431–451 (1990).Article 

Google Scholar 
Guiry, E. J. & Szpak, P. Improved quality control criteria for stable carbon and nitrogen isotope measurements of ancient bone collagen. J. Archaeol. Sci. 132, 105416 (2021).CAS 
Article 

Google Scholar 
Thompson, D. R. & Furness, R. W. Stable-isotope ratios of carbon and nitrogen in feathers indicate seasonal dietary shifts in Northern Fulmars. Auk 112, 493–498 (1995).Article 

Google Scholar 
Carter, H. R. & Sealy, S. G. Historical occurrence of the short-tailed Albatross in British Columbia and Washington. 1841–1958. Wildl. Afield 11, 24–38 (2014).
Google Scholar 
Crockford, S. The Archaeological History of Short-tailed Albatross in British Columbia: A Review and Summary of STAL Skeletal Remains, as Compared to Other Avian Species, Identified from Historic and Prehistoric Midden Deposits. Report on file, Canadian Wildlife Service (2003).Borrmann, R. M., Phillips, R. A., Clay, T. A. & Garthe, S. High foraging site fidelity and spatial segregation among individual great black-backed gulls. J. Avian Biol. 50, e02156 (2019).Article 

Google Scholar 
Wilkinson, B. P., Haynes-Sutton, A. M., Meggs, L. & Jodice, P. G. High spatial fidelity among foraging trips of Masked Boobies from Pedro Cays, Jamaica. PLoS ONE 15, e0231654 (2020).CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Araújo, M. S., Bolnick, D. I. & Layman, C. A. The ecological causes of individual specialisation. Ecol. Lett. 14, 948–958 (2011).PubMed 
Article 
PubMed Central 

Google Scholar 
Grémillet, D. et al. Offshore diplomacy, or how seabirds mitigate intra-specific competition: a case study based on GPS tracking of Cape gannets from neighbouring colonies. Mar. Ecol. Prog. Ser. 268, 265–279 (2004).Article 

Google Scholar 
Irons, D. B. Foraging area fidelity of individual seabirds in relation to tidal cycles and flock feeding. Ecology 79, 647–655 (1998).Article 

Google Scholar 
Piper, W. H. Making habitat selection more “familiar”: a review. Behav. Ecol. Sociobiol. 65, 1329–1351 (2011).Article 

Google Scholar 
Davoren, G. K., Montevecchi, W. A. & Anderson, J. T. Search strategies of a pursuit‐diving marine bird and the persistence of prey patches. Ecol. Monogr. 73, 463–481 (2003).Article 

Google Scholar 
Hazen, E. L. et al. Marine top predators as climate and ecosystem sentinels. Front. Ecol. Environ. 17, 565–574 (2019).Article 

Google Scholar 
Dall, S. R. X., Bell, A. M., Bolnick, D. I. & Ratnieks, F. L. An evolutionary ecology of individual differences. Ecol. Lett. 15, 1189–1198 (2012).PubMed 
PubMed Central 
Article 

Google Scholar 
McAllister, N. M. Avian fauna from the Yuquot excavation in The Yuquot Project, Volume 2 (eds. Folan, W. J. & Dewhirst, J.) 103–174 (National Historic Parks and Sites Branch, 1980).Drucker, P. I. The Northern and Central Nootkan tribes. Bureau of American Ethnology Bulletin 144, 1–480 (1951).
Google Scholar 
Lepofsky, D. & Caldwell, M. Indigenous marine resource management on the Northwest Coast of North America. Ecol. Process 2, 12 (2013).Article 

Google Scholar 
Dewhirst, J. The Indigenous Archaeology of Yuquout, a Nootkan Outside Village (National Historic Parks and Sites Branch, 1980).Longin, R. New method of collagen extraction for radiocarbon dating. Nature 230, 241–242 (1971).CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
Guiry, E. J. & Hunt, B. P. V. Integrating fish scale and bone isotopic compositions for ‘deep time’ retrospective studies. Mar. Environ. Res. 160, 104982 (2020).CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
Hobson, K. A., Atwell, L. & Wassenaar, L. I. Influence of drinking water and diet on the stable-hydrogen isotope ratios of animal tissues. Proc. Natl Acad. Sci. 96, 8003–8006 (1999).CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Qi, H., Coplen, T. B., Geilmann, H., Brand, W. A. & Böhlke, J. K. Two new organic reference materials for δ13C and δ15N measurements and a new value for the δ13C of NBS 22 oil. Rapid Commun. Mass Spectrom 17, 2483–2487 (2003).CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
Qi, H. et al. A new organic reference material, l-glutamic acid, USGS41a, for δ13C and δ15N measurements − a replacement for USGS41. Rapid Commun. Mass Spectrom 30, 859–866 (2016).CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
Szpak, P., Metcalfe, J. Z. & Macdonald, R. A. Best practices for calibrating and reporting stable isotope measurements in archaeology. J. Archaeol. Sci. Rep 13, 609–616 (2017).
Google Scholar 
Hammer, Ø., Haper, A. T. & Ryan, P. D. PAST: paleontological statistics software package for education and data analysis. Palaeontol. Electron. 4, 4 (2001).
Google Scholar 
R Core Team. R: A Language and Environment for Statistical Computing (R Foundation for Statistical Computing, 2019).RStudio Team. RStudio: Integrated Development for R (RStudio, PBC, 2019). More

Experimental setupThis research was approved by the Carnegie Mellon University Committee of the Use of Human Subjects. All methods were performed in accordance with the relevant guidelines and regulations. Informed consent was obtained from all participants. Our data includes no identifying information of human participants. We conducted experiments from February to August 2021 (except for the preliminary sessions of random information; we ran the condition from June to November 2020). We preregistered the main experiment settings using AsPredicted (https://aspredicted.org/sm4k5.pdf).A total of 2786 subjects participated in our incentivized decision-making game experiments. We recruited subjects using Amazon Mechanical Turk (MTurk)52,53. Supplementary Table 1 shows the subject demographics. Our participants interacted anonymously over the Internet using customized software playable in a browser window (available at http://breadboard.yale.edu). All participants provided explicit consent and passed a series of human verification checks and a screening test of understanding game rules and payoffs before playing the game (see SI). We prohibited subjects from participating in more than one session of the experiment by using unique identifications for each subject on MTurk.In each session, subjects were paid a $2.00 show-up fee and a bonus depending on whether they took the appropriate disaster decision in four rounds. Furthermore, subjects earned $1.00 when they completed all four rounds. In each round, when a disaster stroke before they evacuated, the subjects earned no bonus. Otherwise, they earned a bonus of $1.00 without disaster or $0.50 with disaster by spending $0.50 for evacuation, plus $0.05 per other players who took the correct action accordingly (Supplementary Table 2). We have confirmed with prior work that the amount of evacuation cost, if any, makes no significant difference in the game’s performance23.At the start, subjects were required to pass a series of human verification checks. They needed to pass Google’s reCAPTCHA using the “I’m not a robot” checkbox. They were also requested to answer whether they were human players. The exact question asked was: “Please select an applicable answer about you.” The options were: “I am not a bot. I am a real person.” “I am not a real person. I am a bot.” “I am anything but a human.” and “I am a computer program working for a person.” The option’s order was randomized. Only the participants who selected “I am not a bot. I am a real person.” moved to the step of informed consent.When subjects provided explicit consent, they were asked to take a tutorial before the actual game would begin. In the tutorial, each subject separately interacted with three dummy players in two rounds of a 45-s practice game. In the actual game, some subjects would be informed in advance whether a disaster would indeed strike or not. In the practice game, while all subjects were not informed of such information in the first round, they were informed of the information in the second round. Thus, they practiced both conditions in terms of prior information on the disaster (see SI).After the practice game, subjects were assessed for their comprehension of the game rules and payment structure using four multiple-choice questions with three options. If they failed to select the correct answer in one of the questions, they could reselect it only once through the entire test. If they failed to select the correct answer more than once, they were unable to join the actual game.At 720 s after the tutorial beginning, a “Ready” button became visible simultaneously to all the subjects who completed the tutorial and passed the comprehension tests. The actual games started 30 s after the “Ready” button showed up. If subjects did not click the button before the game started, they were dropped. The game required a certain number of subjects. When the subjects who successfully clicked the button were more than 16, surplus subjects, randomly selected, were dropped from the game. When the number of qualified subjects was less than 12, the game did not start. As a result, subjects started the game in a group with an average size of 15.5 (s.d. = 1.1).At the start of the actual game, we selected one subject (the “informant”) at random who was informed in advance whether a disaster would indeed strike or not. The other subjects were informed that some players had accurate information about the disaster, but they were not informed who the informant was. The exact sentence that the informants received in their game screen was “A disaster is going to strike!” when a disaster would strike or “There is no disaster.” when a disaster would not strike. The one that the other uninformed subjects received was “A disaster may or may not strike.” Then, the group had the same informant across the four rounds except for a supplement condition of random informants. In the random informant condition, an informant was randomly selected every round.To prevent an end-of-game effect, we randomly set the game time with a normal distribution of a mean of 75 s and a standard deviation of 10 s. Prior work has confirmed that the game time is sufficient for players to communicate and make an evacuation decision23. As a result, each round ended at 75.0 s on average (s.d. = 9.5) without prior notice. In half of the sessions, a disaster struck at the end of the game. We did not inform any subjects, including the informants, when their sessions would end, the global network structure they were embedded in, or how many informants were in the game. After making their evacuation choice, subjects were informed of their success and failure along with overall results in their group. Then, subjects played another round of the evacuation game until they completed four total rounds. They had the same local network environment across four rounds except for the dynamic network condition.Network structure and tie rewiringIn the network sessions, subjects played the game in a directed network with a random graph configuration. A certain number of ties were present at the game’s onset as the initial density was set to 0.25.In the dynamic network conditions, subjects also could change their neighbors by making or breaking ties between rounds. In the tie-rewiring step, 40% of all the possible subject pairs were chosen at random. Thus, subjects could choose every other player at least once throughout the entire session (i.e., a set of four rounds) with a probability of about 80%. When the chosen pairs were connected, the pairs (the ties) were dissolved if the predecessor subject of the directed ties chose to break the tie. When the chosen pairs were not connected, the pairs (the ties) were newly created when the predecessor of the potential tie chose to create the tie. Subjects were not informed of the rewiring rate.To equalize the game time, we made subjects in the independent and static network conditions wait for additional 10 s after each game round ended. Despite the adjustment, the game time was significantly longer in the dynamic network sessions than in the independent and static network sessions. The average game time is 429.5 s (s.d. = 20.2) for the independent condition; 428.8 s (s.d. = 19.0) for the static network condition; and 564.7 s (s.d. = 36.3) for the dynamic network condition.To clarify mechanisms for dynamic networks to facilitate collective intelligence, we added one supplementary condition. In the supplementary condition, subjects were assigned to one of the 40 isomorphic networks that other subjects had developed with tie-rewiring options through the three rounds in the dynamic network condition (567 subjects in 40 groups). Network structure and other game settings (i.e., whether a disaster stroke, how long the game was, and which node was the informant) were identical to where the others played the game at the final round. However, players were different, and they had no prior experience in the game. They played the game in a network with a topology created by others ostensibly to optimize the accurate flow of information. In contrast to other conditions, subjects played only one round in the isomorphic network condition.Signal buttonsDuring the game of network sessions, subjects were allowed to share information about the possibly impending “disaster” by using “Safe” and “Danger” buttons that indicated their assessment (see SI). The default node color was grey. Then, when they clicked the Safe button, their node turned blue and, after 5 s, automatically returned to grey. Likewise, the Danger button turned their node to red for 5 s. Subjects could see only the colors of neighbors to whom they were directly connected. Since the signal exchange occurred through directed connections, an individual could send, but not receive, information from another subject (and vice versa). Once subjects chose to evacuate, they could no longer send signals, and their node showed grey (the default color) for the rest of the game. The neighbors of evacuated subjects were not informed of their evacuation. We have confirmed with prior work that collective performance does not vary with the communication continuity and the evacuation visibility23. Subjects could use the Safe and Danger buttons any time unless they evacuated, or they did not have to.Players dropping during the gameAfter each game round, when a player was inactive for 10 s, they were warned about being dropped. When they remained inactive after 10 s, they were dropped. When the selected informant was dropped, the session stopped at the round, and we did not use the data. Furthermore, as too many dropped players could affect the network structure and the behavioral dynamics of remaining players, we did not use the sessions where more than 25% of initial players were dropped during the game. Overall, 4 players dropped in 15 sessions; 3 players dropped in 22 sessions; 2 players dropped in 41 sessions; 1 player dropped in 44 sessions; and no player dropped in 58 sessions. The dropped players were prohibited from joining another session of this experiment.As noted above, players took the additional tie-rewiring step every round in the dynamic network sessions. Thus, the total game time was longer in the dynamic network sessions than in the independent and static network sessions even with the adjustment. As a result, more players were dropped in the dynamic network sessions than in the independent and static network sessions. The average number of dropped players across the four rounds is 0.40 (s.d. = 0.60) for the independent condition; 1.15 (s.d. = 0.86) for the static network condition; 1.75 (s.d. = 1.19) for the dynamic network condition. Although group size could affect collective performance, we found the differences in group size small enough for our study. We have confirmed the dynamic network’s performance improvement with a comprehensive analysis controlling the effect of group size (Supplementary Table 3). Also, there was no statistically significant difference in the dropped players’ performance of the dynamic network condition, compared with the other two conditions. The rate of correct actions of dropped players is 0.456 (s.d. = 0.322) for the independent condition, 0.594 (s.d. = 0.387) for the static network condition, and 0.558 (s.d. = 0.411) for the dynamic network condition; P = 0.106 between the independent condition and the dynamic network condition; P = 0.599 between the static network condition and the dynamic network condition (Welch two-sample t test).Analysis of signal diffusionsTo examine the change in signal diffusion, we analyzed “diffusion chains” for each signal type in the network sessions. We first identified the subjects who sent a signal when their neighbors had never sent one as spontaneous “diffusion sources.” When a subject sent a signal after at least one neighbor had sent the same type of signal, we regarded the subject’s signaling (and evacuation with danger signals) as occurring in a chain of signal diffusion and the total number of the responded subjects (including the diffusion source) as the diffusion size.We analyzed the distribution of signal diffusion chains with complementary cumulative distribution functions, measuring the fraction of diffusion chains that exhibit a given number of diffusion sizes. We found that the number of diffusions of both signals did not change across rounds. Safe-signal diffusions were more likely to occur than danger-signal diffusions regardless of whether a “disaster” would strike and how many rounds subjects played. On the other hand, the diffusion size varied greatly across rounds in disaster situations. With “disaster,” false safe signals spread further than true danger signals at the first round, but after that, warnings outperformed safe signals in terms of diffusion size. Figure 2B and Supplementary Fig. 3 scrutinize the changes in diffusion chains with their distributions.Analysis of individual responsivenessWe analyzed how individual evacuation behavior varies with exposure to signals from neighbors54. Let$${a}_{i}^{evacuate},, (t)=left{begin{array}{ll}1&quad text{if subject } i text{ evacuates at time } t\ 0&quad text{otherwise}end{array}right.$$$${a}_{i}^{show, safe},, (t)=left{begin{array}{ll}1&quad text{if subject } i mathrm{ shows a safe signal at time } t\ 0&quad text{otherwise}end{array}right.$$$${a}_{i}^{show , danger} ,, (t)=left{begin{array}{ll}1&quad text{if subject } i text{ shows a danger signal at time } t\ 0&quad text{otherwise}end{array}right.$$The hazard function, or instantaneous rate of occurrence of subject (i)’s evacuation at time t, is defined as:$${lambda }_{i},, (t)=underset{mathit{dt}to 0}{{mathrm{lim}}}frac{{mathrm{Pr}}({a}_{i}^{evacuate}=1;,, tt)}{dt}$$To model the time to evacuation, We used a Cox proportional hazards model with time-varying covariates for the number of signals, incorporating an individual actor-specific random effect55:$${lambda }_{i} ,, left{t|{{P}_{i}, X}_{i}(t), {G}_{i},{Y}_{i}(t)right}={lambda }_{0}(t)mathrm{exp}left{{{beta }_{P}^{{prime}}{P}_{i}+beta }_{X}^{{prime}}{X}_{i}(t)+{beta }_{G}^{{prime}}{G}_{i}+{beta }_{Y}^{{prime}}{Y}_{i}(t)+{gamma }_{i}right}$$where λ0(t) is a baseline hazard at time t.In the model, the hazard λi(t) depends on the covariates Pi, Xi(t), Gi, and Yi(t). The covariate Pi is the vector of subject i’s experiences before the sessions; that is, the number of rounds, the number of disasters that she has experienced, and the number of disasters that she has been struck by.The covariate Xi(t) is the vector of the number of safe signals ({x}_{i}^{safe} (t)), the number of danger signals ({x}_{i}^{danger} (t)). When subject j is a neighbor of subject i (i.e., (jin {N}_{i})), subject i is exposed to the signal of subject j, so that:$${x}_{i}^{safe},, (t)=sum_{jin {N}_{i}}{a}_{j}^{show, safe}(t)$$$${x}_{i}^{danger},, (t)=sum_{jin {N}_{i}}{a}_{j}^{show, danger}(t)$$The covariate Gi is the vector of the properties of the network in which subject i is embedded, out-degree, in-degree, and a network plasticity indicator. The covariate Yi(t) is the vector of the number of the subject i’s actions of sending safe and danger signals before time t. The coefficients β are the fixed effects and γi is the random effect for individual i. We assumed that waiting times to evacuation in different actors are conditionally independent given the sequence of signals they receive from network neighbors. This model shows how the hazard of an individual’s evacuation depends on the signaling actions of others, their network position, and experience (Supplementary Table 4). We applied the same model to the first signaling behavior. More