Ecology

Download PDF

When Mariuá, a 1.5-year-old female jaguar, set foot in our breeding centre in Argentina in December 2018, we did not know that she would make history. Two years later, she walked out with two cubs: the first jaguars to roam the 1.4 million hectares of the Iberá wetlands of northeastern Argentina for at least 70 years. Mariuá and her cubs have started to reverse a process that some had thought irreversible.Within decades, one million species out of a total of some eight million could go extinct globally1. Hunting, habitat loss and ecosystem degradation are propelling this unprecedented biodiversity crisis. Current extinction rates are 100 to 1,000 times higher than in the past several million years.Argentina is no exception. Over the past 150 years, 5 bird and 4 mammal species have gone extinct. Today, about 17% of the country’s 3,000 vertebrate species are imperilled2, and 13 out of the 18 extant species of large mammal, from anteaters to tapirs, are experiencing catastrophic declines, in terms of both number and geographical range (see http://cma.sarem.org.ar).In 1998, we started a rewilding programme in Argentina to try to reverse this appalling loss. Our non-profit foundation, Fundación Rewilding Argentina, was spun out from the US non-profit organization Tompkins Conservation. We create protected areas where we can reintroduce native species, re-establish their interactions, restore ecosystem functionality and build valuable ecotourism based on wildlife viewing.Both rewilding and ecotourism can be controversial. We think that our work is an instructive example of how active restoration of crucial species, when done responsibly, can benefit both ecosystems and local people. It should be in the toolkit for meeting the 2030 biodiversity targets that will be discussed at the Convention on Biological Diversity’s Conference of the Parties in Kunming, China, next month.Three stepsThe popularity of rewilding projects is growing. These include: wolves brought back to Yellowstone National Park in Wyoming, beavers to England, bison and musk ox to northern Russia, leopards to Mozambique and Tasmanian devils to mainland Australia. The International Union for Conservation of Nature reports that, since 2008, at least 418 reintroduction projects have been started3. Most of these projects occur in protected areas and involve one or a few species. Our work in Argentina is broader.As a first step, we acquire private lands with philanthropic funds, reintroduce many species and form government-protected areas that are donated to federal and provincial governments. So far, we have purchased and donated about 400,000 hectares, with an estimated market value of US$91 million. This has created and enlarged six national parks, one national reserve and two provincial parks. Another 100,000 hectares are being donated. Together, these lands comprise a little over 10% of the total terrestrial area currently managed by the National Parks Administration of Argentina.The second step is to restore ecosystems, mainly by reintroducing species at an unprecedented scale. We spend more than $3 million each year on rewilding activities in three regions: the Iberá wetlands in the northeast, the dry Chaco forests in the north and the Patagonian steppe and coast in the south. Most often, we work with species deemed to have large impacts at the ecosystem level, such as large predators and herbivores.

Jaguars now roam Argentina’s Iberá wetlands for the first time in more than 70 years.Credit: Matías Rebak

Thus far, we have successfully reintroduced pampas deer, giant anteaters and collared peccaries (a pig-like, hoofed animal). We have also started founding populations of jaguars, coypus (large aquatic rodents), Wolffsohn’s viscachas (rodents that resemble a large chinchilla), red-and-green macaws and bare-faced curassows (birds related to chickens and pheasants). We are currently working on the reintroduction of 14 species.As they become abundant, reintroduced species re-weave the fabric of ecological relationships. For example, jaguars (Panthera onca) and macaws (Ara chloropterus) are reviving a crucial interaction: predation. Jaguars have begun to prey on eight species, including native rodents and feral hogs, which could limit those populations and thus benefit vegetation growth. The macaws are consuming 49 plant species, which could enhance seed dispersal, although this remains to be tested.
Include the true value of nature when rebuilding economies after coronavirus
Third, we invest heavily in infrastructure, capacity building and publicity to create an economy based on ecotourism. The species we work with are often highly charismatic, which benefits local communities, creating an economic incentive to conserve native wildlife and habitats. We organize workshops and courses so that locals can train as nature guides, cooks, craftspeople and more. In Iberá, where our work is most advanced, tourist visits increased by 87% between 2015 and 2021, according to official data from the Iberá wetland management agency. There were more than 50,000 visitors last year, despite the COVID-19 pandemic.All of these steps are important: simply setting aside protected areas is not enough. Globally, most modern ecosystems are ecologically damaged4, even in long-standing protected areas5. In Argentina, for example, functional populations of jaguars are missing from 19 of 22 national parks where historical distribution data suggest this key apex predator should occur.Jaguars and capybarasOur flagship project is the rewilding of the Iberá wetland. There, we are working on the restoration of nine species, including jaguars, which were eradicated from this area more than 70 years ago. We have now established a founding population of eight individuals: one adult male and three adult females, two of which (including Mariuá) were each released with two cubs aged four months. Our goal is to release a total of 20 individuals by 2027.Of all the species we work with, giant otters (Pteronura brasiliensis) and macaws have been the most difficult. Both species are extinct in the wild in Argentina. Bureaucratic hurdles have made sourcing wild individuals from neighbouring countries impossible.We obtained two pairs of giant otters from European zoos, and are holding them in pens in the core of Iberá. After several attempts, one pair bred successfully and the female gave birth to three cubs, producing the first litter born in the country for more than 30 years. We plan to release this family to the wild next year.

This female giant river otter, together with a male and their three cubs, will be released to the wild in Argentina next year to create a founding population.Credit: Matías Rebak

We source macaws, which have been extinct in the wild in Argentina for 100 years, from zoos, wildlife shelters and breeding centres. Because of their captive origin, we must give them the opportunity to practise flying in an aviary. We provide them with native foods, so that they learn what to eat, and we use a remote-controlled stuffed fox to teach them to avoid predators. This training isn’t always successful. Out of the 87 macaws that we have worked with, 48 were healthy and skilled enough to release. Two founding populations now thrive in the wild; one of them began reproducing in 2020.Efforts elsewhere have demonstrated the powerful effects of restoring species. In the northeast Pacific Ocean, reintroduced sea otters (Enhydra lutris) have voraciously eaten sea urchins, which in turn has allowed the return of lush kelp forests6. In Yellowstone Park, some researchers argue that reintroduced wolves have discouraged herbivores from foraging along stream edges, which might have increased tree growth and stabilized stream banks7. In Mozambique’s Gorongosa Park, the return of wildebeest and other large herbivores has curtailed Mimosa pigra, an undesirable invasive shrub8.
Biodiversity needs every tool in the box: use OECMs
Our rewilding work in Argentina could also have profound impacts. Close monitoring of the female jaguars and their cubs in the Iberá wetland has shown that they are largely feeding on the most abundant native prey: capybaras (Hydrochoerus hydrochaeris). Reducing the number of capybaras is expected to allow more vegetation to thrive, providing habitat for arthropods and small vertebrates, and possibly increasing carbon sequestration9. It could also help to reduce the transmission of sarcoptic mange, a density-dependent disease plaguing the capybara population. Jaguars also prey on foxes, which might benefit threatened bird species. We are working with several academic institutions to test how the return of the jaguar is reshaping the ecosystem.Challenges and caveatsAs our rewilding work gained momentum, critics ramped up from different fronts. At first, some were fearful of our policy of acquiring private lands with funds provided largely by foreign philanthropists. Those concerns faded when we began donating the land to federal and provincial governments.Then, ranchers argued that we were taking agricultural land out of production and reintroducing or boosting populations of animals that would conflict with their livestock. For example, in Patagonia, we established several protected areas where pumas (Puma concolor) and guanacos (Lama guanicoe, a relative of the llama) thrive. For almost a century, ranchers have trapped, shot and poisoned these animals, blaming them for killing sheep and competing for forage, respectively. We are conducting research to quantify the impact of pumas and guanacos on livestock, and offering alternative job opportunities based on wildlife viewing.

Red-and-green macaws went extinct in Argentina in the late 1800s. Rewilding efforts that began in 2016 have now established two founding populations in the Iberá wetlands.Credit: Matías Rebak

Federal and state managers, and often academics, argue that some founding populations of reintroduced species are too small and genetically related to create a viable, long-term population. This is true in some cases. But careful releases of unrelated animals can sidestep this issue. Worries about the spread of diseases when translocating individuals is also often invoked as a reason to halt rewilding activities. We implement thorough health checks and rigorous quarantines to decrease the risk of introducing unwanted diseases in the regions where we work.Concerns are sometimes raised about whether reintroduced species will recreate historical conditions, or instead create something new. Rewilding, however, seeks to regenerate and maintain ecological processes and biodiversity, rather than reaching some specific, historical equilibrium10. We think it is preferable to assume the uncertainties in trying to restore ecosystems, rather than accepting their degraded state.
Protect the last of the wild
Another worry is the possible impacts that tourism can have on climate, biodiversity and society — for instance, on water use, aviation emissions, road building and so on. Our strategy is to limit visitor numbers and avoid crowding by constructing multiple access gates on existing dirt roads.There are many policies that hinder rather than help rewilding. In Argentina, the laws that regulate transportation of wildlife species are built on the assumption that such activities always represent a threat to conservation. Wild animals can typically be imported to the country only through an airport in Buenos Aires. Because of this, an animal that could be driven in a truck from Brazil in a few hours must instead fly more than 1,500 kilometres and then be driven all the way back to its release area. Receiving wild animals at another international port, or moving them around within the country, requires special permits that often take months to obtain. Regulations could be altered to ease rewilding efforts while still policing the illegal wildlife trade.Next stepsNature-based tourism has been growing globally at rates of more than 4% per year, particularly in low- and middle-income countries11. Charismatic fauna, including large predators, are becoming increasingly important. In the Brazilian Pantanal, the world’s largest wetland, wildlife viewing — mostly of jaguars — generated an annual revenue of $6.8 million in 2015. This is three times the revenue obtained from traditional cattle ranching in that region12.With about 97% of the planet’s land surface ravaged by humans4, nature is facing its last stand. Urgent measures are needed not only to halt but also to reverse ecosystem and biodiversity loss. The active reintroduction of key species is one powerful way to heal some degraded ecosystems.This daunting task should not fall solely to non-profit organizations that have limited funds and staff, like us. The United Nations launched its Decade on Ecosystem Restoration in June 2021, calling for massive restoration efforts worldwide to heal nature and the climate. To achieve meaningful results at a global scale, rewilding needs the support of many stakeholders and effective international cooperation. Crucially, it requires the active involvement of governments to facilitate, fund and lead restoration efforts.

Nature 603, 225-227 (2022)
doi: https://doi.org/10.1038/d41586-022-00631-4

ReferencesIPBES. Summary for Policymakers of the Global Assessment Report on Biodiversity and Ecosystem Services of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services (eds Diaz, S. et al.) (IPBES, 2019).
Google Scholar 
Bauni, V. et al. Zookeys 1085, 101–127 (2022).Article 

Google Scholar 
Soorae, P. S. (ed.) Global Conservation Translocation Perspectives: 2021. Case Studies from Around the Globe (IUCN SSC Conservation Translocation Specialist Group, Environment Agency — Abu Dhabi & Calgary Zoo, 2021).
Google Scholar 
Plumptre, A. J. et al. Front. For. Glob. Chang. https://doi.org/10.3389/ffgc.2021.626635 (2021).Article 

Google Scholar 
Jones, K. R. et al. Science 360, 788–791 (2018).PubMed 
Article 

Google Scholar 
Estes, J. A. & Duggins, D. O. Ecol. Monogr. 65, 75–100 (1995).Article 

Google Scholar 
Allen, B. L. et al. Food Webs 12, 64–75 (2017).Article 

Google Scholar 
Guyton, J. A. et al. Nature Ecol. Evol. 4, 712–724 (2020).PubMed 
Article 

Google Scholar 
Schmitz, O. J. et al. Science 362, eaar3213 (2018).PubMed 
Article 

Google Scholar 
Svenning, J. C., Munk, M. & Schweiger, A. in Rewilding (eds Pettorelli, N., Durant, S. M. & Du Toit, J. T.) 73–98 (Cambridge Univ. Press, 2019).
Google Scholar 
Balmford, A. et al. PLoS Biol. 7, e1000144 (2009).PubMed 
Article 

Google Scholar 
Tortato, F. R., Izzo, T. J., Hoogesteijn, R. & Peres, C. A. Glob. Ecol. Conserv. 11, 106–114 (2017).Article 

Google Scholar 
Download references

Competing Interests
The authors declare no competing interests.

Related Articles

Biodiversity needs every tool in the box: use OECMs

Collection: Global biodiversity policy post-2020

Protect the last of the wild

Include the true value of nature when rebuilding economies after coronavirus

Georgina Mace (1953–2020)

Subjects

Ecology

Conservation biology

Biodiversity

Latest on:

Ecology

How itchy vicuñas remade a vast wilderness
Research Highlight 04 MAR 22

How colonialism fed the flames of Australia’s catastrophic wildfires
Research Highlight 24 FEB 22

Apply Singapore Index on Cities’ Biodiversity at scale
Correspondence 22 FEB 22

Biodiversity

Apply Singapore Index on Cities’ Biodiversity at scale
Correspondence 22 FEB 22

Do not downplay biodiversity loss
Matters Arising 26 JAN 22

Shifting baselines and biodiversity success stories
Matters Arising 26 JAN 22

Jobs

Senior Scientific Officer – Functional Genomics (Dr Rachael Natrajan)

Institute of Cancer Research (ICR)
London, United Kingdom

Senior Laboratory Research Scientist

Francis Crick Institute
London, United Kingdom

Postdoctoral Training Fellow – Proteomics

Francis Crick Institute
London, United Kingdom

Laboratory Operations Lead / Hub Lead as Science Support Manager

Francis Crick Institute
London, United Kingdom More

ACIA (2004) Impacts of a Warming Arctic: Arctic Climate Impact Assessment. Cambridge University Press, Cambridge, UK
Google Scholar 
Alsos IG, Ehrich D, Thuiller W, Eidesen PB, Tribsch A, Schonswetter P et al. (2012) Genetic consequences of climate change for northern plants. Proc R Soc B-Biol Sci 279:2042–2051
Google Scholar 
Andrews KR, Good JM, Miller MR, Luikart G, Hohenlohe PA (2016) Harnessing the power of RADseq for ecological and evolutionary genomics. Nat Rev Genet 17:81–92CAS 
PubMed 
PubMed Central 

Google Scholar 
Archer FI, Adams PE, Schneiders BB (2017) strataG: an R package for manipulating, summarizing and analysing population genetic data. Mol Ecol Resour 17:5–11CAS 
PubMed 

Google Scholar 
Arenas M, Ray N, Currat M, Excoffier L (2011) Consequences of range contractions and range shifts on molecular diversity. Mol Biol Evol 29:207–218PubMed 

Google Scholar 
Beaumont MA (1999) Detecting population expansion and decline using microsatellites. Genetics 153:2013–2029CAS 
PubMed 
PubMed Central 

Google Scholar 
Benjamini Y, Hochberg Y (1995). Controlling the false discovery rate: a practical and powerful approach to multiple testing. J Royal Statistical Soc Series B 57:289–300BirdLife International (2018). Pagophila eburnea. The IUCN Red List of Threatened Species 2018: e.T22694473A132555020. https://doi.org/10.2305/IUCN.UK.2018-2.RLTS.T22694473A132555020.en. Downloaded on11 June 2020Boertmann D, Petersen IK, Nielsen HH (2020) Ivory Gull population status in Greenland 2019. Dan Orn Foren Tidsskr 114:141–150
Google Scholar 
Boitard S, Rodríguez W, Jay F, Mona S, Austerlitz F (2016) Inferring population size history from large samples of genome-wide molecular data – an approximate Bayesian computation approach. PLOS Genet 12:1–36
Google Scholar 
Box JE, Colgan WT, Christensen TR, Schmidt NM, Lund M, Parmentier F-JW et al. (2019) Key indicators of Arctic climate change: 1971–2017. Environ Res Lett 14:045010CAS 

Google Scholar 
Braune BM, Mallory ML, Gilchrist HG (2006) Elevated mercury levels in a declining population of ivory gulls in the Canadian Arctic. Mar Pollut Bull 52:978–982CAS 
PubMed 

Google Scholar 
Broquet T, Angelone S, Jaquiéry J, Joly P, Léna JP, Lengagne T et al. (2010) Genetic bottlenecks driven by population disconnection. Conserv Biol 24:1596–1605PubMed 

Google Scholar 
Chen IC, Hill JK, Ohlemüller R, Roy DB, Thomas CD (2011) Rapid range shifts of species associated with high levels of climate warming. Science 333:1024–1026CAS 
PubMed 

Google Scholar 
Chikhi L, Sousa VC, Luisi P, Goossens B, Beaumont MA (2010) The confounding effects of population structure, genetic diversity and the sampling scheme on the detection and quantification of population size changes. Genetics 186:983–995PubMed 
PubMed Central 

Google Scholar 
Collevatti RG, Nabout JC, Diniz-Filho JAF (2011) Range shift and loss of genetic diversity under climate change in Caryocar brasiliense, a Neotropical tree species. Tree Genet Genomes 7:1237–1247
Google Scholar 
Cornuet JM, Luikart G (1996) Description and power analysis of two tests for detecting recent population bottlenecks from allele frequency data. Genetics 144:2001–2014CAS 
PubMed 
PubMed Central 

Google Scholar 
Cotto O, Schmid M, Guillaume F (2020) Nemo-age: spatially explicit simulations of eco-evolutionary dynamics in stage-structured populations under changing environments. Methods Ecol Evol 11:1227–1236
Google Scholar 
Cubaynes S, Doherty PF, Schreiber EA, Gimenez O (2011) To breed or not to breed: a seabird’s response to extreme climatic events. Biol Lett 7:303–306PubMed 

Google Scholar 
Di Rienzo A, Peterson AC, Garza JC, Valdes AM, Slatkin M, Freimer NB (1994). Mutational processes of simple-sequence repeat loci in human populations. Proc Natl Acad Sci USA 91: 3166–3170Do C, Waples RS, Peel D, Macbeth GM, Tillett BJ, Ovenden JR (2014) NeEstimator V2: re-implementation of software for the estimation of contemporary effective population size (Ne) from genetic data. Mol Ecol Resour 14:209–214CAS 
PubMed 

Google Scholar 
Ellegren H (2004) Microsatellites: simple sequences with complex evolution. Nat Rev Genet 5:435–445CAS 
PubMed 

Google Scholar 
England PR, Cornuet J-M, Berthier P, Tallmon DA, Luikart G (2006) Estimating effective population size from linkage disequilibrium: severe biasin small samples. Conserv Genet 7:303
Google Scholar 
Engler JO, Secondi J, Dawson DA, Elle O, Hochkirch A (2016) Range expansion and retraction along a moving contact zone has no effect on the genetic diversity of two passerine birds. Ecography 39:884–893
Google Scholar 
Environment Canada (2014) Recovery Strategy for the Ivory Gull (Pagophila eburnea) in Canada. Species at Risk Act Recovery Strategy Series. Environment Canada, Ottawa. iv+ 21 ppEstoup A, Angers B (1998) Microsatellites and minisatellites for molecular ecology: theoretical and empirical considerations. In Carvalho GR (ed) Advances in molecular ecology, 55–85. IOS Press, Burke, Virginia, USAEwens WJ (1972) The sampling theory of selectively neutral alleles. Theor Popul Biol 3:87–112CAS 
PubMed 

Google Scholar 
Felsenstein J (1971) Inbreeding and variance effective numbers in populations with overlapping generations. Genetics 168:581–597
Google Scholar 
Fort J, Moe B, Strøm H, Grémillet D, Welcker J, Schultner J et al. (2013) Multicolony tracking reveals potential threats to little auks wintering in the North Atlantic from marine pollution and shrinking sea ice cover. Diversity Distrib 19:1322–1332
Google Scholar 
Fraïsse C, Popovic I, Mazoyer C, Spataro B, Delmotte S, Romiguier J et al. (2021). DILS: Demographic inferences with linked selection by using ABC. Mol Ecol Resourc 21:2629–2644Frankham R (1995) Effective population size/adult population size ratios in wildlife: a review. Genetical Res 66:95–107
Google Scholar 
Frankham R, Bradshaw CJA, Brook BW (2014) Genetics in conservation management: revised recommendations for the 50/500 rules, Red List criteria and population viability analyses. Biol Conserv 170:56–63
Google Scholar 
Garnier J, Lewis MA (2016) Expansion under climate change: the genetic consequences. Bull Math Biol 78:2165–2185PubMed 

Google Scholar 
Garza JC, Williamson EG (2001) Detection of reduction in population size using data from microsatellite loci. Mol Ecol 10:305–318CAS 
PubMed 

Google Scholar 
Gavrilo M, Martynova D (2017) Conservation of rare species of marine flora and fauna of the Russian Arctic National Park, included in the Red Data Book of the Russian Federation and in the IUCN Red List. Nat Conserv Res 2:10–42
Google Scholar 
Gienapp P, Teplitsky C, Alho JS, Mills JA, Merila J (2008) Climate change and evolution: disentangling environmental and genetic responses. Mol Ecol 17:167–178CAS 
PubMed 

Google Scholar 
Gilchrist HG, Mallory ML (2005) Declines in abundance and distribution of the ivory gull (Pagophila eburnea) in Arctic Canada. Biol Conserv 121:303–309
Google Scholar 
Gilchrist HG, Strøm H, Gavrilo MV, Mosbech A (2008). International ivory gull conservation strategy and action plan. CAFF International Secretariat,Circumpolar Seabird Group (CBird), CAFF Technical Report No. 18Gilg O, Boertmann D, Merkel F, Aebischer A, Sabard B (2009) Status of the endangered ivory gull, Pagophila eburnea, in Greenland. Polar Biol 32:1275–1286
Google Scholar 
Gilg O, Istomina L, Heygster G, Strøm H, Gavrilo M, Mallory ML et al. (2016) Living on the edge of a shrinking habitat: the ivory gull, Pagophila eburnea, an endangered sea-ice specialist. Biol Lett 12:20160277PubMed 
PubMed Central 

Google Scholar 
Gilg O, Kovacs KM, Aars J, Fort J, Gauthier G, Gremillet D et al. (2012) Climate change and the ecology and evolution of Arctic vertebrates. Ann N Y Acad Sci 1249:166–190PubMed 

Google Scholar 
Goutte A, Kriloff M, Weimerskirch H, Chastel O (2011) Why do some adult birds skip breeding? A hormonal investigation in a long-lived bird. Biol Lett 7:790–792PubMed 
PubMed Central 

Google Scholar 
Guillaume F, Rougemont J (2006) Nemo: an evolutionary and population genetics programming framework. Bioinformatics 22:2556–2557CAS 
PubMed 
PubMed Central 

Google Scholar 
Hill WG (1972) Effective size of populations with overlapping generations. Theor Popul Biol 3:278–289CAS 
PubMed 

Google Scholar 
Hoban SM, Mezzavilla M, Gaggiotti OE, Benazzo A, van Oosterhout C, Bertorelle G (2013) High variance in reproductive success generates a false signature of a genetic bottleneck in populations of constant size: a simulation study. BMC Bioinforma 14:309
Google Scholar 
Hoffmann AA, Sgrò CM (2011) Climate change and evolutionary adaptation. Nature 470:479–485CAS 
PubMed 

Google Scholar 
Hohenlohe PA, Funk WC, Rajora OP (2021) Population genomics for wildlife conservation and management. Mol Ecol 30:62–82PubMed 

Google Scholar 
Jamieson IG, Allendorf FW (2012) How does the 50/500 rule apply to MVPs? Trends Ecol Evol 27:578–584PubMed 

Google Scholar 
Kimura M, Ohta T (1978) Stepwise mutation model and distribution of allelic frequencies in a finite population. Proc Natl Acad Sci USA 75:2868–2872CAS 
PubMed 
PubMed Central 

Google Scholar 
Laporte V, Charlesworth B (2002) Effective population size and population subdivision in demographically structured populations. Genetics 162:501–519CAS 
PubMed 
PubMed Central 

Google Scholar 
Leblois R, Pudlo P, Néron J, Bertaux F, Reddy Beeravolu C, Vitalis R et al. (2014) Maximum-Likelihood Inference of Population Size Contractions from Microsatellite Data. Mol Biol Evol 31:2805–2823CAS 
PubMed 

Google Scholar 
Li H, Durbin R (2011) Inference of human population history from individual whole-genome sequences. Nature 475:493–496CAS 
PubMed 
PubMed Central 

Google Scholar 
Liu X, Fu Y-X (2015) Exploring population size changes using SNP frequency spectra. Nat Genet 47:555–559CAS 
PubMed 
PubMed Central 

Google Scholar 
Lucia M, Verboven N, Strom H, Miljeteig C, Gavrilo MV, Braune BM et al. (2015) Circumpolar contamination in eggs of the high-arctic ivory gull Pagophila eburnea. Environ Toxicol Chem 34:1552–1561CAS 
PubMed 

Google Scholar 
Luikart G, Cornuet J-M (1998) Empirical evaluation of a test for identifying recently bottlenecked populations from allele frequency data. Conserv Biol 12:228–237
Google Scholar 
Mallory ML, Allard KA, Braune BM, Gilchrist HG, Thomas VG (2012) New longevity record for ivory gulls (Pagophila eburnea) and evidence of natal philopatry. Arctic 65:98–101
Google Scholar 
Marandel F, Charrier G, Lamy J-B, Le Cam S, Lorance P, Trenkel VM (2020) Estimating effective population size using RADseq: Effects of SNP selection and sample size. Ecol Evol 10:1929–1937PubMed 
PubMed Central 

Google Scholar 
McInerny GJ, Turner JRG, Wong HY, Travis JMJ, Benton TG (2009) How range shifts induced by climate change affect neutral evolution. Proc R Soc B: Biol Sci 276:1527–1534CAS 

Google Scholar 
McRae L, Deinet S, Gill M, Collen B (2012) Arctic species trend index: tracking trends in Arctic marine populations. CAFF Assessment Series No. 7. Conservation of Arctic Flora and Fauna, Iceland
Google Scholar 
Meredith M, Sommerkorn M, Cassotta S, Derksen C, Ekaykin A, Hollowed A et al. (2020) Polar Regions. In: Pörtner HO, Roberts DC, Masson-Delmotte V, Zhai P, Tignor M, Poloczanska E, Mintenbeck K, Alegría A, Nicolai M, Okem A, Petzold J, Rama B, Weyer NM (eds.) IPCC Special Report on the Ocean and Cryosphere in a Changing Climate. Intergovernmental Panel on Climate Change (IPCC)) Geneva, pp 203–320Miljeteig C, Strom H, Gavrilo MV, Volkov A, Jenssen BM, Gabrielsen GW (2009) High levels of contaminants in ivory gull Pagophila eburnea eggs from the Russian and Norwegian Arctic. Environ Sci Technol 43:5521–5528CAS 
PubMed 

Google Scholar 
Miller MP, Haig SM, Mullins TD, Popper KJ, Green M (2012) Evidence for population bottlenecks and subtle genetic structure in the yellow rail. Condor 114:100–112
Google Scholar 
Nadachowska-Brzyska K, Li C, Smeds L, Zhang G, Ellegren H (2015) Temporal dynamics of avian populations during Pleistocene revealed by whole-genome sequences. Curr Biol 25:1375–1380CAS 
PubMed 
PubMed Central 

Google Scholar 
Nunney L (1993) The influence of mating system and overlapping generations on effective population size. Evolution 47:1329–1341PubMed 

Google Scholar 
Nunney L, Elam DR (1994) Estimating the Effective Population Size of Conserved Populations. Conserv Biol 8:175–184
Google Scholar 
Nunziata SO, Weisrock DW (2018) Estimation of contemporary effective population size and population declines using RAD sequence data. Heredity 120:196–207CAS 
PubMed 

Google Scholar 
Nyström V, Angerbjörn A, Dalén L (2006) Genetic consequences of a demographic bottleneck in the Scandinavian arctic fox. Oikos 114:84–94
Google Scholar 
Orive ME (1993) Effective population size in organisms with complex life-histories. Theor Popul Biol 44:316–340CAS 
PubMed 

Google Scholar 
Parmesan C (2006) Ecological and evolutionary responses to recent climate change. Annu Rev Ecol Evol Syst 37:637–669
Google Scholar 
Parreira BR, Chikhi L (2015) On some genetic consequences of social structure, mating systems, dispersal, and sampling. Proc Natl Acad Sci USA 112:E3318CAS 
PubMed 
PubMed Central 

Google Scholar 
Parreira B, Quéméré E, Vanpé C, Carvalho I, Chikhi L (2020) Genetic consequences of social structure in the golden-crowned sifaka. Heredity 125:328–339PubMed 
PubMed Central 

Google Scholar 
Peart CR, Tusso S, Pophaly SD, Botero-Castro F, Wu C-C, Aurioles-Gamboa D et al. (2020) Determinants of genetic variation across eco-evolutionary scales in pinnipeds. Nat Ecol Evol 4:1095–1104PubMed 

Google Scholar 
Peery MZ, Kirby R, Reid BN, Stoelting R, Doucet-Bëer E, Robinson S et al. (2012) Reliability of genetic bottleneck tests for detecting recent population declines. Mol Ecol 21:3403–3418PubMed 

Google Scholar 
Piry S, Luikart G, Cornuet J-M (1999) Computer note. BOTTLENECK: a computer program for detecting recent reductions in the effective size using allele frequency data. J Heredity 90:502–503
Google Scholar 
Raymond M, Rousset F (1995) GENEPOP (version 1.2): population genetics software for exact tests and ecumenicism. J Heredity 86:248–249
Google Scholar 
Rousset F (1999) Genetic Differentiation in Populations with Different Classes of Individuals. Theor Popul Biol 55:297–308CAS 
PubMed 

Google Scholar 
Rousset F (2008) Genepop’007: a complete re-implementation of the Genepop software for windows and linux. Mol Ecol Notes 8:103–1006
Google Scholar 
Rousset F, Beeravolu CR, Leblois R (2018) Likelihood computation and inference of demographic and mutational parameters from population genetic data under coalescent approximations. J de la Société Française de Statistique 159:142–166
Google Scholar 
Rubidge EM, Patton JL, Lim M, Burton AC, Brashares JS, Moritz C (2012) Climate-induced range contraction drives genetic erosion in an alpine mammal. Nat Clim Change 2:285–288
Google Scholar 
Shafer ABA, Gattepaille LM, Stewart REA, Wolf JBW (2015) Demographic inferences using short-read genomic data in an approximate Bayesian computation framework: in silico evaluation of power, biases and proof of concept in Atlantic walrus. Mol Ecol 24:328–345PubMed 

Google Scholar 
Spencer NC, Gilchrist HG, Mallory ML (2014) Annual movement patterns of endangered ivory gulls: the importance of sea ice. Plos ONE 9:e115231PubMed 
PubMed Central 

Google Scholar 
Stenhouse IJ, Robertson GJ, Gilchrist HG (2004) Recoveries and survival rates of ivory gulls (Pagophila eburnea) banded in Nunavut, Canada 1971–1999. Waterbirds 27:486–492
Google Scholar 
Storz J, Ramakrishnan U, Alberts S (2002) Genetic effective size of a wild primate population: influence of current and historical demography. Evolution 56:817–29PubMed 

Google Scholar 
Strøm H, Bakken V, Skoglund, Descamps S, Fjeldheim VB, Steen H (2020) Population status and trend of the threatened ivory gull Pagophila eburnea in Svalbard. Endanger Species Res 43:435–445
Google Scholar 
Volkov AE, de Korte J (2000) Breeding ecology of the Ivory Gull (Pagophila eburnea) in Sedov Archipelago, Severnaya Zemlya. Heritage of the Russian Arctic. Research, conservation and international cooperation. Ecopros Publishers, Moscow, p 483–500
Google Scholar 
Waples RS (2016) Life-history traits and effective population size in species with overlapping generations revisited: the importance of adult mortality. Heredity 117:241–250CAS 
PubMed 
PubMed Central 

Google Scholar 
Waples RS, Antao T, Luikart G (2014) Effects of overlapping generations on linkage disequilibrium estimates of effective population size. Genetics 197:769–780PubMed 
PubMed Central 

Google Scholar 
Waples RS, Do C (2010) Linkage disequilibrium estimates of contemporary Ne using highly variable genetic markers: a largely untapped resource for applied conservation and evolution. Evolut Appl 3:244–262
Google Scholar 
Waples RS, Luikart G, Faulkner JR, Tallmon DA (2013) Simple life-history traits explain key effective population size ratios across diverse taxa. Proc R Soc B: Biol Sci 280:20131339
Google Scholar 
Weir BS, Cockerham CC (1984) Estimating F-statistics for the analysis of population structure. Evolution 38:1358–1370CAS 
PubMed 
PubMed Central 

Google Scholar 
Williamson-Natesan EG (2005) Comparison of methods for detecting bottlenecks from microsatellite loci. Conserv Genet 6:551–562
Google Scholar 
Wogan GOU, Voelker G, Oatley G, Bowie RCK (2020) Biome stability predicts population structure of a southern African aridland bird species. Ecol Evol 10:4066–4081PubMed 
PubMed Central 

Google Scholar 
Xenikoudakis G, Ersmark E, Tison J-L, Waits L, Kindberg J, Swenson JE et al. (2015) Consequences of a demographic bottleneck on genetic structure and variation in the Scandinavian brown bear. Mol Ecol 24:3441–3454CAS 
PubMed 

Google Scholar 
Yannic G, Broquet T, Strøm H, Aebischer A, Dufresnes C, Gavrilo MV et al. (2016) Genetic and morphological sex identification methods reveal a male-biased sex ratio in the Ivory Gull Pagophila eburnea. J Ornithol 157:861–873
Google Scholar 
Yannic G, Sermier R, Aebischer A, Gavrilo MV, Gilg O, Miljeteig C et al. (2011) Description of microsatellite markers and genotyping performances using feathers and buccal swabs for the ivory gull (Pagophila eburnea). Mol Ecol Resour 11:877–889PubMed 

Google Scholar 
Yannic G, Yearsley J, Sermier R, Dufresnes C, Gilg O, Aebischer A et al. (2016) High connectivity in a long-lived high-Arctic seabird, the ivory gull Pagophila eburnea. Polar Biol 39:221–236
Google Scholar 
Yurkowski DJ, Auger-Méthé M, Mallory ML, Wong SNP, Gilchrist G, Derocher AE et al. (2019) Abundance and species diversity hotspots of tracked marine predators across the North American Arctic. Diversity Distrib 25:328–345
Google Scholar  More

Cape gannet movement trackingThe study took place in the Western Cape, South Africa, where we studied chick-rearing Cape gannets from Malgas Island (33.05° S, 17.93° E) during October–November from 2008 to 2015 (Fig. 1). We caught birds using a pole fitted with a loop and fitted 197 adult Cape gannets (22 in 2008, 16 in 2009, 38 in 2010, 11 in 2011, 29 in 2012, 29 in 2013, 23 in 2014, and 29 in 2015) with GPS-loggers (2008: GPS mass 65 g, i.e., 2.4 % adult body mass, Technosmart, Rom. 2009–2010: GPS mass 45 g, i.e., 1.7% adult body mass, Technosmart, Rom. From 2011: GPS mass 30 g, 1.1% of bird body mass, Catnip Technologies, Hong-Kong). Loggers were attached to the lower back with waterproof Tesa® tape and recorded position at a regular 30-s to 2-min intervals, reinterpolated over 1-min intervals. Devices were recovered after one foraging trip lasting a few hours to one week. Bird handling and tracking using these procedures do not have a measurable impact on foraging behavior19,20. We caught adult birds at-random from the colony, and previous studies showed that this resulted in a well-balanced sex-ratio preventing confounding sex effects21. All experiments were performed under permit from South African National Parks with respect to animal ethics (N° RYAP/AGR/001-2002/V1).Cape gannet movement tactics and behavioral phasesWe identified two movement trip tactics for Cape gannets: After their daytime foraging activities, some birds returned to the colony at night (rest at colony tactic) while others spent all the night at sea (rest at sea tactic). Within the GPS tracks of gannets from these two categories, we discriminated resting, foraging, and commuting phases, with a segmentation-clustering method based on smoothed speed (i.e., speed smoothed over two steps before and after the focal location) and turning angle measured at constant step length. This corresponded to the angle between the focal location, the first location entering a circle of radius equal to the median step length, and the last location inside the circle22. We fitted behavioral identification with the segclust2d package23 for the R software24. See complete details on behavioral classification for Cape gannets tracks in Appendix 1 in Courbin et al.25.Cape fur seal movement tracking and the seascape of fearWe assessed the at-sea spatial distribution of Cape fur seals, a predator of Cape gannet fledglings7 and adults (Supplementary Data 1). We used Argos data collected from 25 lactating female seals before (2003 and 2004) and again concomitantly with gannet tracking (2012 and 2014). Seals were tracked during the same period of the year as gannets (i.e., September to November). Adult females nursing pups were selected at random and captured using a modified hoop net. Once restrained, anesthesia was induced using isoflurane gas delivered via a portable vaporizer (Stinger, Advanced Anesthesia Specialists, Gladesville, New South Wales, Australia). A satellite tag was glued to the guard hairs on the upper back. Individuals were allowed to recover from the anesthesia and resumed normal behavior within 45 min of capture. Throughout the process, the animals’ breathing was closely monitored and their flippers were repeatedly flushed with seawater to prevent hyperthermia. Seals were equipped with Argos satellite transmitters at three colonies (Fig. 1): Kleinsee (29°35’09”S, 16°59’56”E) located ~400 km to the North of the gannet colony (n = 8 seals in 2003 and 2004); Vondeling Island (33°09’11”S, 17°58’57”E), ~12 km away from the gannet colony (n = 12 seals in 2012 and 2014); and Geyser Rock (34°41’19”S, 19°24’49”E) located ~230 km to the South of the gannet colony (n = 5 seals in 2003). Seals at Vondeling Island were equipped with Argos-linked Spot-6 position transmitting tags (Wildlife Computers) following deployment procedures outlined in Kirkman et al.26. Seals at Kleinsee and Geyser Rock were equipped with ST18 and ST20 satellite-linked platform terminal transmitters (Telonics, Mesa, USA), as detailed in Skern-Mauritzen et al.27. Devices collected a well-balanced number of Argos locations during the day (n = 6080 locations) and at night (n = 6501 locations). See full details on seal tracking in Supplementary Table 6. All fieldwork was permitted by the Animal Ethics Committee of the Department of Environmental Affairs and Tourism’s Marine and Coastal Management branch, which at the time was the management authority of South Africa’s marine and coastal environment (Ref: DEAT2006-06-23).We modeled both daytime and nighttime at-sea occurrences of seals for each colony with resource selection functions (RSF)28,29, a proxy of the fear effect for Cape gannets. RSF compared environmental features of seal’s at-sea Argos positions (i.e., further 500 m than the colony) with five times more random locations that captured the breadth of environmental conditions available to seals. We sampled random locations for each individual within the yearly area used by seals from each colony, delineated by the 95% kernel utilization distribution of the Argos locations of all seals of the colony. RSF were fitted with a generalized linear mixed model with a binomial distribution for errors. As environmental variables, we considered bathymetry (m), the slope of the bathymetry (°) and the distance to the colony (km) within the RSF. These variables were not highly correlated (|r| ≤ 0.61) and had low collinearity with a variance inflation factor VIF  More

Alberdi, M. T. A review of Old World hipparionine horses in The Evolution of Perissodactyls (eds. Prothero, D. R. & Schoch, R. M.) 234–261 (Clarendon Press, Oxford University Press, New York, NY ⋅ Oxford, 1989).Sen S. Hipparion Datum and its chronologic evidence in the Mediterranean area in European Neogene Mammal Chronology (eds. Lindsay, E. H., Fahlbusch, V. & Mein, P.) 495–506 (Plenum Press, New York, 1990).Garcés, M., Cabrera, L., Agustí, J. & Parés, J. M. Old World first appearance datum of “Hipparion” horses: Late Miocene large–mammal dispersal and global events. Geology 25(1), 19–22. https://doi.org/10.1130/0091-7613(1997)025%3c0019:OWFADO%3e2.3.CO;2 (1997).ADS 
Article 

Google Scholar 
Woodburne, M. O. A new occurrence of Cormohipparion, with implications for the Old World Hippotherium Datum. J. Vert. Paleont. 25(1), 256–257 (2005).Woodburne, M. O. Phyletic diversification of the Cormohipparion occidentale complex (Mammalia; Perissodactyla, Equidae), Late Miocene, North America, and the origin of the Old World Hippotherium Datum. B. Am. Mus. Nat. Hist. 306, 1–180 (2007).Article 

Google Scholar 
Bernor, R. L., Qiu, Z. & Tobien, H., 1987. Phylogenetic and biogeographic bases for an Old World hipparionine horse geochronology. Proceedings of the VIIIth International Congress of the Regional Committee on Mediterranean Neogene Stratigraphy, Budapest. Ann. Inst. Geol. Publ. Hung. 70, 43–53 (1987).Bernor, R. L., Tobien, H., Hayek, L–A. C. & Mittmann, H. –W. Hippotherium primigenium (Equidae, Mammalia) from the late Miocene of Höwenegg (Hegau, Germany). Andrias 10, 1–230 (1997).Qiu, Z., Huang, W. & Guo, Z. The Chinese hipparionine fossils. Palaeont. Sin. New Ser C 25, 1–250 (1987) ((in Chinese with English summary)).
Google Scholar 
Zouhri, S. & Bensalmia, A. Révision systématique des Hipparion sensu lato (Perissodactyla, Equidae) de l’Ancien Monde. Estud. Geol. 61, 61–99. https://doi.org/10.3989/egeol.05611-243 (2005).Article 

Google Scholar 
Liu, Y. Late Miocene hipparionine fossils from Lantian, Shaanxi Province and phylogenetic analysis on Chinese Hipparionines. PhD thesis (University of Chinese Academy of Sciences, Beijing) (2013).Bernor, R. L., Wang, S., Liu, Y., Chen, Y. & Sun, B. Shanxihippus dermatorhinus (new gen.) with comparisons to Old World hipparions with specialized nasal apparati. Riv. Ital. Paleontol. Stratigr. 124, 361–386 (2018).Bernor, R. L., Boaz, N. T., Omar, C., El-Shawaihdi, M. H. & Rook, L. Sahabi Eurygnathohippus feibeli: Its systematic, stratigraphic, chronologic and biogeographic contexts. Riv. Ital. Paleontol. Stratigr. 126, 561–581 (2020).
Google Scholar 
Liu, T., Li, C. & Zhai, R. Pliocene mammalian fauna of Lantian, Shaangxi. Prof. Pap. Stratigr. Paleont. 7, 149–200 (1978) (in Chinese).Deng, T. et al. Locomotive implication of a Pliocene three–toed horse skeleton from Tibet and its paleo–altimetry significance. Proc. Natl. Acad. Sci. USA 109, 7374–7378. https://doi.org/10.1073/pnas.1201052109 (2012).ADS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Fang, X., Garzione, C., van der Voo, R., Li, J. & Fan, M. Flexural subsidence by 29 Ma on the NE edge of Tibet from the magnetostratigraphy of Linxia Basin China. Earth Planet. Sci. Lett. 210, 545–560 (2003).ADS 
CAS 
Article 

Google Scholar 
Fang, X. et al. Tectonosedimentary evolution model of an intracontinental flexural (foreland) basin for paleoclimatic research. Glob. Planet Change 145, 78–97. https://doi.org/10.1016/j.gloplacha.2016.08.015 (2016).ADS 
Article 

Google Scholar 
Deng, T., Qiu, Z., Wang, B., Wang, X. & Hou, S. Chapter 9: Late Cenozoic Biostratigraphy of the Linxia Basin, Northwestern China in Fossil Mammals of Asia: Neogene Biostratigraphy and Chronology (eds. Wang, X., Flynn, L. J. & Fortelius, M.) 243–273 (Columbia University Press, New York, 2013).Li, Y., Deng, T., Hua, H., Li, Y. & Zhang, Y. Assessment of dental ontogeny in late Miocene hipparionines from the Lamagou fauna of Fugu, Shaanxi Province China. PLoS ONE https://doi.org/10.1371/journal.pone.0175460 (2017).Article 
PubMed 
PubMed Central 

Google Scholar 
Li, Y., Deng, T., Hua, H., Sun, B. & Zhang, Y. Locomotor adaptations of 7.4 Ma Hipparionine fossils from the middle reaches of the Yellow River and their palaeoecological significance. Hist. Biol. 33(7), 927–940 (2021).Bernor, R. L. & Sun, B. Morphology through ontogeny of Chinese Proboscidipparion and Plesiohipparion and observations on their Eurasian and African relatives. Vert. PalAsiat. 53, 73–92 (2015).
Google Scholar 
Woodburne, M. O. & Bernor, R. L. On superspecific groups of some Old World hipparinonine horses. J. Paleontol. 54(6), 1319–1348 (1980).
Google Scholar 
Bernor, R. L., Kaya, F., Kaakinen, A., Saarinen, J. & Fortelius, M. Old World hipparion evolution, biogeography, climatology and ecology. Earth Sci. Rev. 211, 103784 (2021).Article 

Google Scholar 
Bernor, R. L., Göhlich, U. B., Harzhauser, M. & Semprebon, G. M. The Pannonian C hipparions from the Vienna Basin. Palaegeogr. Palaeoclim. Palaeoecol. 476, 28–41. https://doi.org/10.1016/j.palaeo.2017.03.026 (2017).ADS 
Article 

Google Scholar 
Arambourg, C. Vertebres continentaux du Miocene superieur de l’Afrique du Nord. Service Carte Geologie Algerie Paleontologie Memoire, Nouveaux Serie 4, 1–159 (1959).
Google Scholar 
Bernor, R.L. & White, T.D. Systematics and biogeography of “Cormohipparion” africanum, Early Vallesian (MN 9, ca. 10.5 Ma) of Bou Hanifia, Algeria in Papers on Geology, Vertebrate Paleontology, and Biostratigraphy in Honorof Michael O. Woodburne. (ed. Albright, B.), Bull., Mus. of No. Arizona. 65, 635–658 (2009).Bernor, R. L., Scott, R. S., Fortelius, M., Kappelman, J. & Sen, S. Systematics and Evolution of the late Miocene Hipparions from Sinap, Turkey in The Geology and Paleontology of the Miocene Sinap Formation, Turkey (eds. Fortelius, M., Kappelman, J., Sen, S. & Bernor, R. L.) 220–281 (Columbia University Press, New York, 2003).Sack, W. O. The stay–apparatus of the horse’s hindlimb, explained. Equine Pract. 11, 31–35 (1988).
Google Scholar 
MacFadden, B. J. In Fossil Horses: Systematics, Paleobiology, and Evolution of the Family Equidae (Cambridge University Press, 1992).
Google Scholar 
Li, F. & Li D. Latest Miocene Hipparion (Plesiohipparion) of Zanda Basin in Paleontology of the Ngari Area, Tibet (Xizang) (eds. Yang, Z. & Nie, Z.) 186–193 (China University of Geosciences Press, Wuhan, 1990).Thomason, J. J. The functional morphology of the manus in tridactyl equids Merychippus and Mesohippus: Paleontological inferences from neontological models. J. Vert. Paleont. 6, 143–161 (1986).Article 

Google Scholar 
Eisenmann, V. What metapodial morphometry has to say about some Miocene Hipparions in Paleoclimate and Evolution, with Emphasis on Human Origins (eds. Vrba, E. S., Denton, G. H., Partridge, T. C. & Burckle, L. H.) 148–164 (Yale University Press, New Haven, 1995).Deng, T. & Wang, X. Late Miocene Hipparion (Equidae, Mammalia) of eastern Qaidam Basin in Qinghai, China. Vert. PalAsiat. 42(4), 316–333 (2004) (in Chinese with English summary).Wang, X. et al. Vertebrate paleontology, biostratigraphy, geochronology, and paleoenvironment of Qaidam Basin in northern Tibetan Plateau. Palaegeogr. Palaeoclim. Palaeoecol. 254, 363–385 (2007).Article 

Google Scholar 
Zhang, Z. et al. Chapter 6: Mammalian Biochronology of the Late Miocene Bahe Formation in Fossil Mammals of Asia: Neogene Biostratigraphy and Chronology (eds. Wang, X., Flynn, L. J. & Fortelius, M.) 187–202 (Columbia University Press, New York, 2013).Xue, X. X., Zhang, Y. X. & Yue, L. P. Discovery of Hipparion fauna of Laogaochuan and its division of eras, Fugu County Shaanxi. Chin. Sci. Bull. 40, 447–449 (1995).Article 

Google Scholar 
Deng, T. Late Cenozoic environmental changes in the Linxia Basin (Gansu, China) as indicated by cenograms of fossil mammals. Vert. PalAsiat. 47(4), 282–298 (2009).
Google Scholar 
An, Z., Kutzbach, J. E., Prell, W. L. & Porter, S. C. Evolution of Asian monsoons and phased uplift of the Himalaya-Tibetan plateau since Late Miocene times. Nature 411, 62–66. https://doi.org/10.1038/35075035 (2001).ADS 
CAS 
Article 

Google Scholar 
An, Z. et al. Changes of the monsoon–arid environment in China and growth of the Tibetan Plateau since the Miocene. Q. Sci. 26(5), 678–693. https://doi.org/10.3321/j.issn:1001-7410.2006.05.002 (2006).Article 

Google Scholar 
Wang, X. et al. Origin of the Red Earth sequence on the northeastern Tibetan Plateau and its implications for regional aridity since the middle Miocene. Sci. China D Earth Sci. 49(5), 505–517. https://doi.org/10.1007/s11430-006-0505-3 (2006).ADS 
CAS 
Article 

Google Scholar 
Dettman, D. L., Fang, X., Garzione, C. N. & Li, J. Uplift–driven climate change at 12 Ma: A long δ18O record from the NE margin of the Tibetan plateau. Earth Planet. Sci. Lett. 214, 267–277. https://doi.org/10.1016/S0012-821X(03)00383-2 (2003).ADS 
CAS 
Article 

Google Scholar 
Jiang, H. C. & Ding, Z. L. A 20 Ma pollen record of East-Asian summer monsoon evolution from Guyuan, Ningxia China. Palaegeogr. Palaeoclim. Palaeoecol. 265, 30–38. https://doi.org/10.1016/j.palaeo.2008.04.016 (2008).ADS 
Article 

Google Scholar 
Fortelius, M. et al. Late Miocene and Pliocene large land mammals and climatic changes in Eurasia. Palaegeogr. Palaeoclim. Palaeoecol. 238, 219–227. https://doi.org/10.1016/j.palaeo.2006.03.042 (2006).ADS 
Article 

Google Scholar 
Fortelius, M. et al. Evolution of neogene mammals in Eurasia: Environmental forcing and biotic interactions. Annu. Rev. Earth Pl Sci. 42, 579–604 (2014).ADS 
CAS 
Article 

Google Scholar 
Bernor, R. L., Scott, R. S., Fortelius, M., Kappelman, J. & Sen, S. Equidae (Perissodactyla) in The Geology and Paleontology of the Miocene Sinap Formation, Turkey. (eds. Fortelius, M., Kappelman, J., Sen, S. & Bernor, R.L.) 220–281 (Columbia University Press, New York, 2003).Zhang, Z. S., Wang, H. J., Guo, Z. T. & Jiang, D. B. What triggers the transition of palaeoenvironmental patterns in China, the Tibetan Plateau uplift or the Paratethys Sea retreat?. Palaegeogr. Palaeoclim. Palaeoecol. 245, 317–331. https://doi.org/10.1016/j.palaeo.2006.08.003 (2007).ADS 
Article 

Google Scholar 
Böhme, M., Ilg, A. & Winklhofer, M. Late Miocene “washhouse” climate in Europe. Earth Planet. Sci. Lett. 275, 393–401. https://doi.org/10.1016/j.epsl.2008.09.011 (2008).ADS 
CAS 
Article 

Google Scholar 
Janis, C. M., Damuth, J. & Theodor, J. M. Miocene ungulates and terrestrial primary productivity: Where have all the browsers gone?. Proc. Natl. Acad. Sci. USA 97(14), 7899–7904. https://doi.org/10.1073/pnas.97.14.7899 (2000).ADS 
CAS 
Article 
PubMed 

Google Scholar 
Janis, C. M., Damuth, J. & Theodor, J. M. The origins and evolution of the North American grassland biome: The story from the hoofed mammals. Palaegeogr. Palaeoclim. Palaeoecol. 177, 183–198. https://doi.org/10.1016/S0031-0182(01)00359-5 (2002).ADS 
Article 

Google Scholar 
Mihlbachler, M. C., Rivals, F., Solounias, N. & Semprebon, G. M. Dietary change and evolution of horses in North America. Science 331, 1178–1181. https://doi.org/10.1126/science.1196166 (2011).ADS 
CAS 
Article 
PubMed 

Google Scholar 
Hayek, L. C., Bernor, R. L., Solounias, N. & Steigerwald, A. Preliminary studies of hipparionine horse diet as measured by tooth microwear. Ann. Zool. Fenniuci 28, 187–200 (1992).
Google Scholar 
Sisson, S. The anatomy of the domestic animals (Saunders W B Comp, 1953).
Google Scholar 
Budras, K.-D., Sack, W. O. & Röck, S. In Anatomy of the horse. (Schlütersche, Hannover, 2009).Eisenmann, V., Alberdi, M. T., de Giuli, C. & Staesche, U. In Studying Fossil Horses, Vol. I: Methodology. (ed. Brill, E. J.) 1–71 (Leiden, 1988).Deng, T., Hou, S. & Wang, S. Neogene integrative stratigraphy and timescale of China. Sci. China Earth Sci. 62, 310–323 (2019).ADS 
CAS 
Article 

Google Scholar  More

Witman, J. D. & Smit, F. Rapid community change at a tropical upwelling site in the Galapagos Marine Reserve. Biodivers. Conserv. 12, 25–45 (2003).
Google Scholar 
Edgar, G. J., Banks, S., Fariña, J. M., Calvopiña, M. & Martínez, C. Regional biogeography of shallow reef fish and macro-invertebrate communities in the Galapagos archipelago. J. Biogeogr. 31, 1107–1124 (2004).
Google Scholar 
Okey, T. A. et al. A trophic model of a Galápagos subtidal rocky reef for evaluating fisheries and conservation strategies. Ecol. Model. 172, 383–401 (2004).
Google Scholar 
Briggs, J. C. & Bowen, B. W. A realignment of marine biogeographic provinces with particular reference to fish distributions. J. Biogeogr. 39, 12–30 (2012).
Google Scholar 
Salinas de León, P. et al. Largest global shark biomass found in the northern Galápagos Islands of Darwin and Wolf. PeerJ 4, e1911. https://doi.org/10.7717/peerj.1911 (2016).CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Humann, P. & DeLoach, N. Reef Fish Identification: Galápagos (ed. Humann, P.) (New World Publications, Inc., 2003).McCosker, J. E. & Rosenblatt, R. H. The fishes of the Galápagos archipelago: An update. Proc. Calif. Acad. Sci. 61, 167–195 (2010).
Google Scholar 
Grove, J. S. & Lavenberg, R. J. The Fishes of the Galapagos Islands (Stanford University Press, 1997).Allen, G. & Ross-Robertson, D. Fishes of Tropical Eastern Pacific (University of Hawaii Press, 1994).Ruttenberg, B. I., Haupt, A. J., Chiriboga, A. I. & Warner, R. R. Patterns, causes and consequences of regional variation in the ecology and life history of a reef fish. Oecologia 145, 394–403 (2005).ADS 
PubMed 

Google Scholar 
Bernardi, G. et al. Darwin’s fishes: Phylogeography of Galápagos Islands reef fishes. Bull. Mar. Sci. 90, 533–549 (2014).
Google Scholar 
Banks, S., Vera, M. & Chiriboga, A. Establishing reference points to assess long-term change in zooxanthellate coral communities of the northern Galápagos coral reefs. Galapagos Res. 66, 43–64 (2009).
Google Scholar 
Palacios, D., Bograd, S., Foley, D. & Schwing, F. Oceanographic characteristics of biological hot spots in the North Pacific: A remote sensing perspective. Deep Sea Res Part II Top. Stud. Oceanogr. 53, 250–269 (2006).ADS 

Google Scholar 
Sweet, W. V. et al. Water mass seasonal variability in the Galapagos Archipelago. Deep Sea Res. Part I Oceanogr. Res. Pap. 54, 2023–2035 (2007).ADS 

Google Scholar 
Schaeffer, B. et al. Phytoplankton biomass distribution and identification of productive habitats within the Galapagos Marine Reserve by MODIS, a surface acquisition system, and in-situ measurements. Remote Sens. Environ. 112, 3044–3054 (2008).ADS 

Google Scholar 
Witman, J. D., Brandt, M. & Smith, F. Coupling between subtidal prey and consumers along a mesoscale upwelling gradient in the Galapagos Islands. Ecol. Monogr. 80, 153–177 (2010).
Google Scholar 
Moity, N. Evaluation of no-take zones in the Galápagos marine reserve, zoning plan 2000. Frontiers. 5, 244. https://doi.org/10.3389/fmars.2018.00244 (2018).Article 

Google Scholar 
Lamb, R. W., Smith, F. & Witman, J. D. Consumer mobility predicts impacts of herbivory across an environmental stress gradient. Ecology 101, e02910. https://doi.org/10.1002/ecy.2910 (2020).Article 
PubMed 

Google Scholar 
Edgar, G. J. et al. Conservation of threatened species in the Galapagos Marine Reserve through identification and protection of marine key biodiversity areas. Aquat. Conserv. 18, 955–968 (2008).
Google Scholar 
Carrión-Cortez, J. A., Zárate, P. & Seminoff, J. A. Feeding ecology of the green sea turtle (Chelonia mydas) in the Galapagos Islands. J. Mar. Biol. Assoc. U. K. 90, 1005–1013 (2010).
Google Scholar 
Moity, N., Delgado, B. & Salinas-de-León, P. Correction: Mangroves in the Galapagos islands: Distribution and dynamics. PLoS One 14, e0212440. https://doi.org/10.1371/journal.pone.0212440 (2019).Article 
PubMed 
PubMed Central 

Google Scholar 
Seitz, R. D., Wennhage, H., Bergström, U., Lipcius, R. N. & Ysebaert, T. Ecological value of coastal habitats for commercially and ecologically important species. ICES J. Mar. Sci. 71, 648–665 (2014).
Google Scholar 
Aguaiza, C. The role of mangrove as nursery habitats for coral reef fish species in the Galapagos Islands. MSc Thesis (University of Queensland, 2016).Llerena-Martillo, Y., Peñaherrera-Palma, C. & Espinoza, E. Fish assemblages in three fringed mangrove bays of Santa Cruz Island, Galapagos Marine Reserve. Rev. Biol. Trop. 66, 674–687 (2018).
Google Scholar 
Fierro-Arcos, D. et al. Mangrove fish assemblages reflect the environmental diversity of the Galapagos Islands. Mar. Ecol. Prog. Ser. 664, 183–205 (2021).ADS 

Google Scholar 
Henseler, C. et al. Coastal habitats and their importance for the diversity of benthic communities: A species-and trait-based approach. Estuar. Coast. Shelf Sci. 226, 106272. https://doi.org/10.1016/j.ecss.2019.106272 (2019).Article 

Google Scholar 
Loreau, M. et al. Biodiversity and ecosystem functioning: Current knowledge and future challenges. Science 294, 804–808 (2001).ADS 
CAS 

Google Scholar 
Menezes, R. F. et al. Variation in fish community structure, richness, and diversity in 56 Danish lakes with contrasting depth, size, and trophic state: Does the method matter?. Hydrobiologia 710, 47–59 (2013).
Google Scholar 
Hu, M., Wang, C., Liu, Y., Zhang, X. & Jian, S. Fish species composition, distribution and community structure in the lower reaches of Ganjiang River, Jiangxi, China. Sci. Rep. 9, 10100. https://doi.org/10.1038/s41598-019-46600-2 (2019).ADS 
CAS 
Article 
PubMed 

Google Scholar 
Clarke, K. R. & Warwick, R. M. Changes in Marine Communities: An Approach to Statistical Analysis and Interpretation, 2nd ed. (PRIMER-E Ltd, Plymouth Marine Laboratory, 2001).Warwick, R. M. & Clarke, K. R. New biodiversity measures reveal a decrease in taxonomic distinctness with increasing stress. Mar. Ecol. Prog. Ser. 129, 301–305 (1995).ADS 

Google Scholar 
Clarke, K. R. & Warwick, R. M. The taxonomic distinctness measure of biodiversity: Weighting of step lengths between hierarchical levels. Mar. Ecol. Prog. Ser. 184, 21–29 (1999).ADS 

Google Scholar 
Nieto-Navarro, J. T., Zetina-Rejón, M. A., Arreguín-Sánchez, F., Palacios-Salgado, D. & Jordán, F. Changes in fish bycatch during the shrimp fishing season along the eastern coast of the mouth of the Gulf of California. J. Appl. Ichthyol. 29, 610–616 (2013).
Google Scholar 
Escobar-Toledo, F., Zetina-Rejón, M. J. & Duarte, L. O. Measuring the spatial and seasonal variability of community structure and diversity of fish by-catch from tropical shrimp trawling in the Colombian Caribbean Sea. Mar. Biol. Res. 11, 528–539 (2015).
Google Scholar 
Herrera-Valdivia, E., López-Martínez, J., Castillo Vargasmachuca, S. & García-Juárez, A. R. Diversidad taxonómica y funcional en la comunidad de peces de la pesca de arrastre de camarón en el norte del Golfo de California, México. Rev. Biol. Trop. 64, 587–602 (2016).PubMed 

Google Scholar 
Heylings, P., Bensted-Smith, R. & Altamirano, M. Zonificación e historia de la Reserva Marina de Galápagos. In Reserva Marina de Galápagos. Línea Base de la Biodiversidad (eds. Danulat, E. & Edgar, G. J.) 10–21 (Fundación Charles Darwin y Servicio Parque Nacional de Galápagos, 2002).Edgar, G. J. et al. Bias in evaluating the effects of marine protected areas: The importance of baseline data for the Galapagos Marine Reserve. Environ. Conserv. 3, 212–218. https://doi.org/10.1017/S0376892904001584 (2004).Article 

Google Scholar 
Jennings, S., Brierley, A. S. & Walker, J. W. The inshore fish assemblages of the Galápagos archipelago. Biol. Conserv. 70, 49–57 (1994).
Google Scholar 
Brito, A., Pérez-Ruzafaga, A. & Bacallado, J. J. Ictiofauna costera de las islas Galápagos: composición y estructura del poblamiento de los fondos rocosos. Res. Cient. Proy. Galápagos TFCM 5, 61 (1997).
Google Scholar 
Bruneel, S. et al. Assessing the drivers behind the structure and diversity of fish assemblages associated with rocky shores in the Galapagos Archipelago. J. Mar. Sci. Eng. 9, 375. https://doi.org/10.3390/jmse9040375 (2021).Article 

Google Scholar 
Wellington, G. M., Strong, A. E. & Merlen, G. Sea surface temperature variation in the Galápagos Archipelago: A comparison between AVHRR nighttime satellite data and in-situ instrumentation (1982–1998). Bull. Mar. Res. 69, 27–42 (2001).
Google Scholar 
Snell, H., Stone, P. & Snell, H. L. A summary of geographical characteristics of the Galapagos Islands. J. Biogeogr. 23, 619–624 (1996).
Google Scholar 
Bustamante, R. H., et al. Outstanding marine features of Galápagos. In A Biodiversity Vision for the Galapagos Islands: An Exercise for Ecoregional Planning (eds. Bensted-Smith, R. & Dinnerstein, E.) 60–71 (WWF, 2002).Airoldi, L. & Beck, M. W. Loss, status and trends for coastal marine habitats of Europe. In Oceanography and Marine Biology: An Annual Review (eds. Gibson, R. N., Atkinson, R. J. A. & Gordon, J. D. M.) vol. 45, 345–405 (Taylor & Francis, 2007).Carr, M. H., Malone, D. P., Hixon, M. A., Holbrook, S. J. & Schmitt, R. J. How Scuba changed our understanding of nature: underwater breakthrough in reef fish ecology. In Research and Discoveries: The Revolution of Science Through Scuba vol. 39, 157–167 (Smithsonian Contributions to the Marine Sciences, 2013).Durkacz, S. Assessing the Oceanographic Conditions and Distribution of Reef Fish Assemblages Throughout the Galápagos Islands Using Underwater Visual Survey Methods. MSc Thesis (Texas A & M University, 2014).Fischer, W. et al. Guía FAO para la identificación de especies para los fines de pesca. Pacífico Centro-Oriental vol. II–III, 648–1652 (FAO, 1995).Clarke, K. R. & Warwick, R. M. A taxonomic distinctness index and its statistical properties. J. Appl. Ecol. 35, 523–531 (1998).
Google Scholar 
Clarke, K. R. Non-parametric multivariate analyses of changes in community structure. Aust. J. Ecol. 18, 117–143 (1993).
Google Scholar 
Rosenberg, A., Binford, T. E., Leathery, S., Hill, R. L. & Bickers, K. Ecosystem approaches to fishery management through essential fish habitat. Bull. Mar. Sci. 66, 535–542 (2000).
Google Scholar 
Aburto-Oropeza, O. & Balart, E. F. Community structure of reef fish in several habitats of a rocky reef in the Gulf of California. Mar. Ecol. 22, 283–305 (2001).ADS 

Google Scholar 
Fulton, C. J., Bellwood, D. R. & Wainwright, P. C. Wave energy and swimming performance shape coral reef fish assemblages. Proc. R. Soc. B 272, 827–832 (2005).CAS 
PubMed 

Google Scholar 
Dominici-Arosemena, A. & Wolff, M. Reef fish community structure in the Tropical Eastern Pacific (Panamá): Living on a relatively stable rocky reef environment. Helgol. Mar. Res. 60, 287–305 (2006).ADS 

Google Scholar 
Villegas-Sánchez, C. A., Abitia-Cárdenas, L. A., Gutiérrez-Sánchez, F. J. & Galván-Magaña, F. Rocky-reef fish assemblages at San José Island, Mexico. Rev. Mex. Biodivers. 80, 169–179 (2009).
Google Scholar 
Wiens, J. J. & Graham, C. H. Niche conservatism: Integrating evolution, ecology, and conservation biology. Annu. Rev. Ecol. Evol. Syst. 36, 519–539 (2005).
Google Scholar 
Glynn, P. Some physical and biological determinants of coral community structure in the eastern Pacific. Ecol. Monogr. 46, 431–456 (1976).
Google Scholar 
Ramos-Miranda, J. et al. Changes in four complementary facets of fish diversity in a tropical coastal lagoon after 18 years: A functional interpretation. Mar. Ecol. Prog. Ser. 304, 1–13 (2005).ADS 

Google Scholar 
Gristina, M., Bahri, T., Fiorentino, F. & Garofalo, G. Comparison of demersal fish assemblages in three areas of the Strait of Sicily under different trawling pressure. Fish. Res. 81, 60–71 (2006).
Google Scholar 
Pérez-Ruzafa, A. P., Marcos, C. & Bacallado, J. J. Biodiversidad marina en archipiélagos e islas: patrones de riqueza específica y afinidades faunísticas. Vieraea Folia Scientarum Biologicarum Canariensium. 33, 455–476 (2005).
Google Scholar 
Malcolm, H. A., Jordan, A. & Smith, S. D. Biogeographical and cross-shelf patterns of reef fish assemblages in a transition zone. Mar. Biodivers. 40(3), 181–193 (2010).
Google Scholar 
García-Charton, J. A. & Pérez-Ruzafa, A. P. Correlation between habitat structure and a rocky reef fish assemblage in the Southwest Mediterranean. Mar. Ecol. 19(2), 111–128 (1998).ADS 

Google Scholar 
Mumby, P. J. et al. Mangroves enhance the biomass of coral reef fish communities in the Caribbean. Nature 427, 533–536 (2004).ADS 
CAS 
PubMed 

Google Scholar 
Unsworth, R. K. F. et al. High connectivity of Indo-Pacific seagrass fish assemblages with mangrove and coral reef habitats. Mar. Ecol. Prog. Ser. 353, 213–224 (2008).ADS 

Google Scholar 
Birkeland, C. & Amesbury, S. S. Fish-transect surveys to determine the influence of neighboring habitats on fish community structure in the tropical Pacific. Co-operation for environmental protection in the Pacific. UNEP Reg. Seas Rep. Stud. 97, 195–202 (1988).
Google Scholar 
Thollot, P., Kulbicki, M., & Wantiez, L. Temporal patterns of species composition in three habitats of the St Vincent Bay area (New Caledonia): Coral reefs, soft bottoms and mangroves. In Proceedings International Soc. Reef Studies. 127–137 (1991).Kulbicki, M. Present knowledge of the structure of coral reef fish assemblages in the Pacific. UNEP Reg. Seas Rep. Stud. 147, 31–53 (1992).
Google Scholar 
Cruz-Romero, M., Chávez, E.A., Espino, E. & García, A. Assessment of a snapper complex (Lutjanus spp.) of the eastern tropical Pacific. In Biology, Fisheries and Culture of Tropical Groupers and Snappers (eds. Arreguín-Sánchez, F., Munro, J. L., Balgos, M. C. & Pauly, D.) 324–330 (ICLARM Conf. Proc. 48, 1996).Aguilar-Santana, F. Biología reproductiva de Prionurus laticlavius (Valenciennes, 1846) (Teleostei: Acanthuridae) en la Costa Sudoccidental del Golfo de California, México. PhD Thesis (Instituto Politécnico Nacional, 2020).Hall, S. The Effects of Fishing on Marine Ecosystems and Communities (Blackwell Science Ltd., 1999).Mangi, S. C. & Roberts, C. M. Quantifying the environmental impacts of artisanal fishing gear on Kenya’s coral reef ecosystems. Mar. Pollut. Bull. 52, 1646–1660 (2006).CAS 
PubMed 

Google Scholar 
Rees, M. J., Jordan, A., Price, O. F., Coleman, M. A. & Davis, A. R. Abiotic surrogates for temperate rocky reef biodiversity: Implications for marine protected areas. Divers. Distrib. 20(3), 284–296 (2014).
Google Scholar 
Ferrari, R. et al. Habitat structural complexity metrics improve predictions of fish abundance and distribution. Ecography 41(7), 1077–1091 (2018).
Google Scholar 
Pihl, L. & Wennhage, H. Structure and diversity of fish assemblages on rocky and soft bottom shores on the Swedish west coast. J. Fish Biol. 61, 148–166 (2002).
Google Scholar 
La Mesa, G., Molinari, A., Gambaccini, S. & Tunesi, L. Spatial pattern of coastal fish assemblages in different habitats in North-western Mediterranean. Mar. Ecol. 32, 104–114 (2011).ADS 

Google Scholar 
Kristensen, L. D. et al. Establishment of blue mussel beds to enhance fish habitats. Appl. Ecol. Environ. Res. 13, 783–798 (2015).
Google Scholar 
Bergström, L., Karlsson, M., Bergström, U., Pihl, L. & Kraufvelin, P. Distribution of mesopredatory fish determined by habitat variables in a predator-depleted coastal system. Mar. Biol. 163, 201. https://doi.org/10.1007/s00227-016-2977-9 (2016).Article 
PubMed 
PubMed Central 

Google Scholar 
Galván-Villa, C. M., Arreola-Robles, J. L., Ríos-Jara, E. & Rodríguez-Zaragoza, F. A. Ensamblajes de peces arrecifales y su relación con el hábitat bentónico de la Isla Isabel, Nayarit, México. Rev. Biol. Mar. Oceanogr. 45, 311–324 (2010).
Google Scholar 
Lunt, J. & Smee, D. L. Turbidity alters estuarine biodiversity and species composition. ICES J. Mar. Sci. 77, 379–387 (2019).
Google Scholar 
Anthony, K. R., Ridd, P. V., Orpin, A. R., Larcombe, P. & Lough, J. Temporal variation of light availability in coastal benthic habitats: Effects of clouds, turbidity, and tides. Limnol. Oceanogr. 49, 2201–2211 (2004).ADS 

Google Scholar 
Helfman, G. S. Patterns of community structure in fishes: Summary and overview’. Environ. Biol. Fishes 3, 129–148 (1978).
Google Scholar 
Helfman, G. S. Fish behaviour by day, night and twilight. In The Behaviour of Teleost Fishes (ed. Pitcher T.J.) (Springer, 1986).Warwick, R. M. & Clarke, K. R. Taxonomic distinctness and environmental assessment. J. Appl. Ecol. 35, 532–543 (1998).
Google Scholar 
Rogers, S. I., Clarke, K. R. & Reynolds, J. D. The taxonomic distinctness of coastal bottom-dwelling fish communities of the North-east Atlantic. J. Anim. Ecol. 68, 769–782 (1999).
Google Scholar 
Robertson, A. I., & Blaber, S. J. M. Plankton, epibenthos and fish communities. In Tropical Mangrove Ecosystems (eds. Robertson, A. I. & Alongi, D. M.) Coastal and Estuarine Studies No. 41, 173–224 (American Geophysical Union, 1992).Koranteng, K. A. Diversity and stability of demersal species assemblages in the Gulf of Guinea. West Afr. J. Appl. Ecol. 2, 49–63 (2001).
Google Scholar 
McCormick, M. I. Comparison of field methods for measuring surface topography and their associations with a tropical reef fish assemblage. Mar. Ecol. Prog. Ser. 112, 87–96 (1994).ADS 

Google Scholar 
Moraes, L. E., Paes, E., Garcia, A., Möller, O. Jr. & Vieira, J. Delayed response of fish abundance to environmental changes: A novel multivariate time-lag approach. Mar. Ecol. Prog. Ser. 456, 159–168 (2012).ADS 

Google Scholar 
Edgar, G. J. et al. El Niño, grazers and fisheries interact to greatly elevate extinction risk for Galapagos marine species. Glob. Change. Biol. 16, 2876–2890 (2010).ADS 

Google Scholar 
Glynn, P. W., Enochs, I. C., Afflerbach, J. A., Brandtneris, V. W. & Serafy, J. E. Eastern Pacific reef fish responses to coral recovery following El Niño disturbances. Mar. Ecol. Prog. Ser. 495, 233–247 (2014).ADS 

Google Scholar  More

Field experiments: collective structuresWe found that self-assembled Eciton hamatum bridges adaptively adjust in response to shifts in the terrain on which they are built. Detailed methods are included in Methods: Field experiments. Briefly, we moved foraging trails onto an apparatus where we could introduce a terrain gap. We repeatedly changed the size of this gap by first incrementally increasing it to 30 mm, by 1 mm every 30 s, and then incrementally contracting it at the same rate (See Fig. 1, Methods: Field experiments, and Supplementary Movie 1). As the size of the gap was expanded (the period before the dotted line in Fig. 2a, b) both the volume and number of ants increased to mean maximum values of 1080 mm3 (standard error, s.e. 84) and 18.9 ants (s.e. 1.6), respectively. Ants typically began forming a bridge when the gap was ~5 mm. As the gap size was decreased (period after dotted line), volume and the number of ants decreased back to zero as ants left the bridge. These broad dynamics across the ten complete trials were similar (Fig. 2a, b, inset panels and Supplementary Figs. 2, 3). Additionally, bridge volume (Fig. 2a) strongly correlated with the number of ants in the bridge (Fig. 2b), indicating that the density of ants per unit volume in these structures is relatively consistent (Pearson correlation coefficients range from 0.88 to 0.98 across the ten trials, see also Supplementary Fig. 4). Bridges broke and quickly reformed in eight of the ten trials; breaks occurred in both experimental phases, and these broken periods were excluded from analyses. Overall, these results show that bridges adjust dynamically to changing terrain geometry, as stretching the bridges caused them to become larger, with more ants, and contracting bridges caused them to become smaller, with fewer ants.Fig. 1: Experimental procedure and data extraction summary.Experiments were conducted on robust E. hamatum foraging trails, which were moved onto the experimental apparatus while it was closed. a Experimental procedure: The size of the gap was increased by 1 mm every 30 s until the gap reached 30 mm (expansion phase), then decreased at the same rate till no gap remained (the contraction phase). b Field setup: Experiments were recorded from both the side and the top, examples of bridges during each phase of the same trial are shown. c Data extraction: Example images and silhouettes from the maximum size bridge (30 mm) of the same trial as the images of 20 mm bridges shown in panel a. The envelopes of the bridges were extracted at a temporal resolution of 1 s; for each focal second, image frames were averaged over 10 s to remove ants walking on the bridge from the extracted envelopes. Envelopes were automatically extracted using hue-saturation-value (HSV) thresholding, with thresholds checked independently for each trial due to lighting differences. Locations of fixed points on the platform were used to re-scale and combine data from the side and top views into a single coordinate system in which 100 pixels = 1 cm. Estimates of bridge volume, mean cross-sectional area, and relative height of the center of mass were recorded from the extracted envelopes as shown. See Methods: Data extraction and Supplementary Note 1 for additional details of the data extraction process, including additional bridge metrics.Full size imageFig. 2: Changes in collective structures in experiments.a, b Volume and group size of self-assembled bridges: a Estimated volume of collective bridge structures over time for one focal trial (main figure) and three other examples (inset). The dotted vertical line indicates the time when the experiment shifted from the expansion phase (increasing gap size) to the contraction phase (decreasing gap size). Gray shading indicates that the bridge was broken or recovering from a break; result metrics may be inaccurate during these periods and they were, therefore, excluded from analyses. b The number of ants in the bridge structure over time for the same focal trial (main figure) and three other examples (inset). c–f Hysteresis: Trials consistently show hysteresis, with bridge status at a particular gap size differing during the expansion and contraction phases, for volume (c), number of ants (d), mean cross-sectional area (e), and tautness, or the height of the center of mass of the bridge from the side view (f; lower values indicate bridge is hanging lower). c–f Panels show result metrics over gap size for the same focal trial as in panels a and b, as well as for three other examples (inset). Points show individual measurements, taken every second, lines are smoothed LOESS (local regression) for the expansion (orange points, dashed orange line) and contraction (green points, solid green line) phases. The area between the smoothed lines (shaded gray) shows the extent of hysteresis. c, e, f) Points are jittered to improve clarity. a–f See Supplementary Figs. 2, 3, 5–8 for all complete trials.Full size imageHowever, these changes were not symmetric—adjustments in the contraction phase were not the inverse of adjustments in the expansion phase. We found consistent hysteresis in several metrics; for a given gap size, bridges were larger and made up of more individuals during the contraction of the gap than the expansion (Fig. 2c, d; t-test for volume: mean extent of hysteresis = 0.43, 95% CI = 0.29 to 0.58, t = 6.7, df = 9, p  More

Model descriptionThe model involves a discrete time, discrete valued stochastic Markov process. Model variables and parameters are given in Tables 1 and 2 respectively. Both time and the number of individuals of each type are constrained to be integer valued. Death and reproductive mutation are stochastic processes derived from sampling from binomial distributions given by the relevant probabilities. All ensemble results give the 100-replicate average for the parameter choices in question.Growth of individuals from species ({S}_{1}) and ({S}_{2}) is proportional to the bio-available level of environmental substances ({R}_{1}) and ({R}_{2}) respectively. At time (t) (where time is in units of biological generations) the change in the number ({N}_{q,j}) of individuals of genotype (j) (non-producer, producer, plastic) within species (q) (({S}_{1}) or ({S}_{2})) can be written as a function of the state of the variables at the previous time-step:$${N}_{q,j}left(t+1right)=left(left({N}_{q,j}left(tright)-{rho }_{q,j}left(tright)right)cdot {G}_{q,j}left(tright)-{{{{{{rm{{Upsilon }}}}}}}}_{q,j}left(tright)-{delta }_{q,j}left(tright)+{{{{{{rm{{Upsilon }}}}}}}}_{q,xne j}left(tright)right)cdot left(1-frac{{S}_{q}left(tright)}{K}right)$$
(1)
The leftmost bracket on the right-hand side represents the number of individuals escaping starvation (death due to insufficient environmental substance) at the previous time-step and ({G}_{q,j}left(tright)) is the per capita reproductive growth rate. ({{{{{{rm{{Upsilon }}}}}}}}_{q,j}left(tright)) gives the number of mutant offspring individuals produced during reproduction from parent individuals of genotype (j). ({{{{{{rm{{Upsilon }}}}}}}}_{q,xne j}left(tright)) represents the number of (j) genotype individuals derived from mutation in parent individuals of other genotypes. ({delta }_{q,j}left(tright)) is the number of individuals of genotype (j) lost to random death, and the rightmost bracket relates the total number ({S}_{q}left(tright)) of individuals of species (q) to carrying capacity (K), which represents limitation of growth by any factor other than the relevant environmental substance, e.g. space. (The steady state population size in all simulations shown is below (K) and limited by the environmental substance influx. The carrying capacity is included in the model for computational reasons and as a “crash preventer” but has no qualitative effect on the results).The total number of individuals ({S}_{q}left(tright)) in species (q) is the sum of the number of individuals of each genotype (producer, non-producer and plastic, as discussed in the main text):$${S}_{q}left(tright)=mathop{sum }limits_{j=1}^{{j}_{{total}}}{N}_{q,j}left(tright)={N}_{q,{prod}}left(tright)+{N}_{q,{non}-{prod}}left(tright)+{N}_{q,{plast}}left(tright)$$
(2)
The genotype-specific reproductive growth rate ({G}_{q,j}left(tright)) (again for genotype (j) within species (q), time (t)), gives the number of offspring individuals produced per parent individual, per time-step. Growth rate is an increasing function of the bio-available level of environmental substance ({R}_{q,{BIOAVAIL}{ABLE}}) (the subscript (q) being identical because species ({S}_{1}) and ({S}_{2}) assimilate substances ({R}_{1}) and ({R}_{2}) respectively). Growth rate also includes a substance-to-biomass conversion efficiency parameter ({f}_{{conv}}) and a genotype-specific per capita term ({G}_{q,j,{PR}}) (number of offspring per parent, per unit environmental substance assimilated, per unit time). In the absence of growth-limitation by environmental substance levels, growth rate is capped at a genotype-specific maximum ({G}_{q,{jMAX}}):$${G}_{q,j}left(tright)={MIN}[{G}_{q,j,{PR}}cdot {R}_{q,{BIOAVAILABLE}}(t)cdot {f}_{{conv}},{G}_{q,{jMAX}}]$$
(3)
$${G}_{q,{non}-{prod},{PR}}={G}_{0}$$
(4)
$${G}_{q,{non}-{prod},{MAX}}={G}_{0}cdot {R}_{{assimMAX}}$$
(5)
({G}_{0}) is the baseline number of offspring, per parent, per unit substance assimilated. ({R}_{{assimMAX}}) is a universal maximum potential number of units of environmental substance that can be assimilated by a single individual per time-step (i.e. representing basic physiological constraints on growth). The producer genotype incurs a per capita reproductive growth rate cost ({kappa }_{{prod}}) relative to the non-producer:$${G}_{q,{prod},{PR}}={G}_{0}cdot (1-{kappa }_{{prod}})$$
(6)
$${G}_{q,{prod},{MAX}}={G}_{0}cdot (1-{kappa }_{{prod}})cdot {R}_{{assimMAX}}$$
(7)
This growth rate formulation is therefore a highly simplified linearization of the Michaelis-Menten kinetics normally used in models of resource and nutrient assimilation.The plastic genotype switches phenotype depending upon the level of environmental substance relative to a fixed threshold ({{R}_{q,{BIOAVAILABLE}}}_{{crit}}), in effect becoming a second non-producer genotype below this threshold and a second producer genotype above it:$${IF}[{R}_{q,{BIOAVAILABLE}}(t)ge {{R}_{q,{BIOAVAILABLE}}}_{{crit}}],{G}_{q,{plast}}left(tright)={G}_{q,{prod}}left(tright)$$$${ELSEIF}[{R}_{q,{BIOAVAILA}{BLE}}left(tright) , < , {{R}_{q,{BIOAVAILABLE}}}_{{crit}}],{G}_{q,{plast}}left(tright)={G}_{q,{non}-{prod}}left(tright)$$ (8) There is no spatial structure whatsoever, thus access to environmental substance is uniform across individuals. The bioavailable quantity of each environmental substance is simply the total amount ({R}_{q,{NET}}(t)) divided by the total number of individuals assimilating it:$${R}_{q,{BIOAVAILABLE}}left(tright)=frac{{R}_{q,{NET}}(t)}{{S}_{q}left(tright)}$$ (9) We allow the per capita reproductive growth rate to fall below ({G}_{q,j}left(tright)=1), which, if interpreted deterministically at the individual level would correspond to an individual failing to sustain its biomass to the next time-step and thus dying. However, a population-level average ({G}_{q,j}left(tright) , < , 1) is interpretable in terms of a thinning factor that maps between discretized individuals and continuously distributed environmental substance. Thus, a thinning factor of (left(1-{G}_{q,j}left(tright)right)) is used to calculate the total number of individuals dying of starvation ({rho }_{q,j}) (again genotype (j), species (q)). This represents pre-reproduction deaths, corresponding to the difference between the actual population size and the population size that the environmental substance pool is capable of supporting. ({rho }_{q,j}left(tright)) is constrained to be an integer and is zero for ({G}_{q,j}left(tright) , > , 1):$${rho }_{q,j}left(tright)={N}_{q,j}left(tright)cdot {MAX}left[0,left(1-{G}_{q,j}left(tright)right)right]$$
(10)
A subset of offspring are a different genotype from their parent via mutation. For parent genotype (j), the number ({{{{{{rm{{Upsilon }}}}}}}}_{q,j}left(tright)) of mutant offspring with genotype (ne j) is calculated using baseline mutation probability per reproductive event ({mu }_{0}), with the total number of new individuals produced by the parent individuals surviving starvation ({G}_{q,j}left(tright)cdot left({N}_{q,j}left(tright)-{rho }_{q,j}left(tright)right)). The total number of mutant offspring ({{{{{{rm{{Upsilon }}}}}}}}_{q,j}left(tright)) is thus a binomially distributed random variable with success probability ({mu }_{0}) and number of trials ({G}_{q,j}left(tright)cdot left({N}_{q,j}left(tright)-{rho }_{q,j}left(tright)right)). The expected value (Eleft[{{{{{{rm{{Upsilon }}}}}}}}_{q,j}left(tright)right]) is the product of these two numbers:$$ {{{{{{rm{{Upsilon }}}}}}}}_{q,j}left(tright) sim Bleft({G}_{q,j}left(tright)cdot left({N}_{q,j}left(tright)-{rho }_{q,j}left(tright)right),{mu }_{0}right), \ Eleft[{{{{{{rm{{Upsilon }}}}}}}}_{q,j}left(tright)right]={G}_{q,j}left(tright)cdot left({N}_{q,j}left(tright)-{rho }_{q,j}left(tright)right)cdot {mu }_{0}$$
(11)
Mutation to genotype (j) from the other genotypes is calculated in exactly the same way using the number and reproductive growth rates of the relevant (other) genotypes. Any particular mutant offspring is randomly allocated to one of the other genotypes with equal probability ({p}_{kto j}=frac{1}{{j}_{{total}}-1}=0.5) (where ({j}_{{total}}=3) is the total number of genotypes per species). The expected number of offspring with genotype (j) produced by mutation within parent offspring of other genotypes (kne j) is therefore:$$E[{{{{{{rm{{Upsilon }}}}}}}}_{q,x , ne , j}left(tright)]={left(mathop{sum }limits_{k=1}^{{k}_{{to}{tal}}}{{{{{{rm{{Upsilon }}}}}}}}_{q,k}left(tright)right)}_{kne j}cdot frac{1}{{j}_{{total}}-1}$$
(12)
Independently of reproduction and assimilation of environmental substance, any given individual has a probability ({delta }_{0}) at each time point of death due to stochastic factors. The genotype/species specific number of such deaths is again a random sample from a binomial distribution, with success probability ({delta }_{0}):$${delta }_{q,j}left(tright) sim Bleft({N}_{q,j}left(tright),{delta }_{0}right),Eleft[{delta }_{q,j}left(tright)right]={N}_{q,j}left(tright)cdot {delta }_{0}$$
(13)
The net quantity of growth-limiting environmental substance at each time-step is given by the difference between total biotic assimilation ({A}_{{R}_{q}}) and the production ({P}_{{R}_{q}}) and abiotic input ({varphi }_{{R}_{q}}) fluxes:$${R}_{q,{NET}}(t+1)={varphi }_{{R}_{q}}(t)+{P}_{{R}_{q}}(t)-{A}_{{R}_{q}}(t)$$
(14)
The abiotic net influx is the sum of two fluxes. First, an input term that is the product of a baseline scaling factor ({{varphi }_{0}}_{{R}_{q}}) and a model forcing (frac{partial {t}_{{geo}}}{partial {t}_{{bio}}}) representing the mapping between abiotic-geological and biotic-evolutionary timescales. In practice (frac{partial {t}_{{geo}}}{partial {t}_{{bio}}}(t)) was set to either (1) or (0) or (in fluctuation runs) a time-dependent switching between the two. (More sophisticated implementations of (frac{partial {t}_{{geo}}}{partial {t}_{{bio}}}(t)), e.g. sinusoidal oscillations and stochastic time dependence, were attempted but made little qualitative difference to the results). Second, an abiotic removal term that scales linearly with the quantity of environmental substance:$${varphi }_{{R}_{q}}(t)={{varphi }_{0}}_{{R}_{q}}cdot frac{partial {t}_{{geo}}}{partial {t}_{{bio}}}left(tright)-frac{{R}_{q,{NET}(t)}}{{R}_{q,{NET}0}}$$
(15)
where ({R}_{q,{NET}0}) is a normalization factor representing the sensitivity of the abiotic efflux to the influx. In the absence of any biota and for (frac{partial {t}_{{geo}}}{partial {t}_{{bio}}}=1) the steady state environmental substance level is immediately given by (15) as ({R}_{q,{NET}(t)}={R}_{q,{NET}0}cdot {{varphi }_{0}}_{{R}_{q}}), thus the numerical value of ({R}_{q,{NET}0}) corresponds to the abiotic steady state residence time.Total biotic assimilation ({A}_{R}) of each environmental substance is given by:$${A}_{{R}_{q}}left(tright)=mathop{sum }limits_{k=1}^{{j}_{{total}}}frac{{G}_{q,k}left(tright)cdot left({N}_{q,k}left(tright)-{rho }_{q,k}left(tright)right)}{{G}_{q,{kPR}}}$$
(16)
The numerator gives the total number of individuals produced as a result of biological assimilation of environmental substance ({R}_{q}) and the denominator is the genotype specific number of individuals produced per unit substance assimilated, dividing through by which therefore converts to total units of substance assimilated by the population as a whole.Net biotic ({P}_{{R}_{q}{NET}}) production of substance ({R}_{q}) by the producer genotype in the other species (p) is calculated equivalently, via the product of the per capita production rate ({P}_{q,{prod}}) and the total number of reproducing individuals:$${P}_{q,{prod}}left(tright)=frac{{MIN}[{G}_{q,{prod}}left(tright)cdot {f}_{{convprod}},{G}_{q,{prodMAX}}]}{{G}_{p,{PRODUCER;PR}}}$$
(17)
$${P}_{{R}_{q}{NET}}left(tright)={P}_{q,{prod}}left(tright)cdot left({N}_{p,{PRODUCER}}left(tright)-{rho }_{p,P{RODUCER}}left(tright)right)$$
(18)
Where ({f}_{{conv},{PROD}}) is the per capita efficiency by which producers convert the environmental substance that they assimilate into the by-product they produce (note that the equivalent conversion efficiency for assimilation ({f}_{{conv}}) already appears in the growth functions of each genotype, therefore does not appear in Eq. (17)).The residence time ({T}_{{R}_{q}})of each environmental substance is given by the net quantity of this substance divided by the influx, to give units of the average number of biological generations a unit of environmental substance spends in the relevant pool before being removed. In those simulations in which the abiotic influx ({varphi }_{{R}_{q}}(t)) was set to zero (i.e. during the shut-off intervals) production ({P}_{{R}_{q}{NE}T}left(tright)) was used as an alternative denominator:$${IF}left[{varphi }_{{R}_{q}}left(tright) , > , 0,{T}_{{R}_{q}}left(tright)=frac{{R}_{q,{NET}}left(tright)}{{varphi }_{{R}_{q}}left(tright)}right]!,{IF}left[left(left({varphi }_{{R}_{q}}left(tright)=0right){& }left({P}_{{R}_{q}{NET}}left(tright) , > ,0right)right)!,{T}_{{R}_{q}}left(tright)=frac{{R}_{q,{NET}}left(tright)}{{P}_{{R}_{q}{NET}}left(tright)}right]{ELSE}[{T}_{{R}_{q}}left(tright)=0]$$
(19)
The cycling ratio ({{CR}}_{{R}_{q}}) of each substance is given by the ratio between net biotic assimilation of that substance ({A}_{{R}_{q}}left(tright)) and the abiotic influx of that substance ({varphi }_{{R}_{q}}left(tright)). As with the residence time, when the abiotic influx was zero, the input from biological production was used as an alternative denominator:$${IF}left[{varphi }_{{R}_{q}}left(tright) , > , 0,{{CR}}_{{R}_{q}}left(tright)=frac{{A}_{{R}_{q}}left(tright)}{{varphi }_{{R}_{q}}left(tright)}right]{IF}left[left(left({varphi }_{{R}_{q}}left(tright)=0right){{& }}left({P}_{{R}_{q}{NET}}left(tright) , > ,0right)right),{{CR}}_{{R}_{q}}left(tright)=frac{{A}_{{R}_{q}}left(tright)}{{P}_{{R}_{q}{NET}}left(tright)}right]{ELSE}[{{CR}}_{{R}_{q}}left(tright)=0]$$
(20)
Deterministic approximation to steady stateAssume that at steady state substance assimilation will reach a maximal state such that the level of environmental substance is limiting to population size. Assume that such a state is below the level ({{R}_{q,{BIOAVAILABLE}}}_{{crit}}) at which the plastic genotype effectively becomes a second non-producer genotype and can thus be subsumed into non-producer frequency, such that (2) becomes ({S}_{q}left(tright)=mathop{sum }nolimits_{j=1}^{{j}_{{total}}}{N}_{q,j}left(tright)={N}_{q,{prod}}left(tright)+{N}_{q,{non}-{prod}}left(tright)). Assume that there are non-zero starvations at each time-step for all genotypes, which implies growth rate ({G}_{q,j}left(tright) , < , 1,ll {G}_{q,{jMAX}},forall j,q), which gives by (3)({G}_{q,j}left(tright)={G}_{q,j,{PR}}cdot {R}_{q,{BIOAVAILABLE}}left(tright)cdot {f}_{{conv}}). Substituting this into (10), then the first bracketed term in (1), then labeling the post-starvation number of individuals as ({({N}_{q,j}left(tright))}_{{NET}}):$${({N}_{q,j}left(tright))}_{{NET}}=left({N}_{q,j}left(tright)-{rho }_{q,j}left(tright)right)={N}_{q,j}left(tright)cdot left(1-left(1-{G}_{q,j}left(tright)right)right)={N}_{q,j}left(tright)cdot {G}_{q,j}left(tright)$$ (21) Approximate (14) deterministically by a fixed fractional parameter corresponding to the baseline random death rate:$${delta }_{q,j}left(tright)approx {N}_{q,j}left(tright)cdot {delta }_{0}$$ (22) Doing the same for mutation:$${{{{{{rm{{Upsilon }}}}}}}}_{q,j}left(tright)={G}_{q,j}left(tright)cdot {left({N}_{q,j}left(tright)right)}_{{NET}}cdot {mu }_{0}={N}_{q,j}left(tright)cdot {{G}_{q,j}left(tright)}^{2}cdot {mu }_{0}$$ (23) The term in (12) for mutation to (j) from other genotypes simplifies to$${{{{{{rm{{Upsilon }}}}}}}}_{q,x ,ne , j}left(tright)={sum }_{k,=,1}^{{j}_{{total}}}frac{{G}_{q,k}left(tright)cdot {left({N}_{q,k}left(tright)right)}_{{NET}}cdot {mu }_{0}}{{j}_{{total}}-1}={N}_{q,k}left(tright)cdot {{G}_{q,k}left(tright)}^{2}cdot {mu }_{0}$$Substituting Eqs. (21–23) into (1):$${N}_{q,j}left(tright)cdot {{G}_{q,j}left(tright)}^{2}cdot (1-{mu }_{0})-{N}_{q,j}left(tright)cdot {delta }_{0}+{N}_{q,k}left(tright)cdot {{G}_{q,k}left(tright)}^{2}cdot {mu }_{0}=0$$ (24) Noting that by Eqs. (3)–(5) combined with the above assumptions, the growth rate of the producer can be written as:$${G}_{q,{prod}}left(tright)={G}_{q,{cheat}}left(tright)cdot (1-{kappa }_{{prod},q})$$ (25) Writing (24) explicitly for each genotype:$${N}_{q,{non}-{prod}}left(tright)cdot {{G}_{q,{non}-{prod}}left(tright)}^{2}cdot (1-{mu }_{0})-{N}_{q,{non}-{prod}}left(tright)cdot {delta }_{0}+{N}_{q,{prod}}left(tright)cdot {{G}_{q,{prod}}left(tright)}^{2}cdot {mu }_{0}=0$$$${N}_{q,{prod}}left(tright)cdot {{G}_{q,{prod}}left(tright)}^{2}cdot (1-{mu }_{0})-{N}_{q,{prod}}left(tright)cdot {delta }_{0}+{N}_{q,{non}-{prod}}left(tright)cdot {{G}_{q,{non}-{prod}}left(tright)}^{2}cdot {mu }_{0}=0$$Adding:$${N}_{q,{non}-{prod}}left(tright)cdot {{G}_{q,{non}-{prod}}left(tright)}^{2}cdot left(1-{mu }_{0}right)-{N}_{q,{non}-{prod}}left(tright)cdot {delta }_{0}$$$$+{N}_{q,{prod}}left(tright)cdot {left({G}_{q,{non}-{prod}}left(tright)cdot left(1-{kappa }_{{prod},q}right)right)}^{2}cdot {mu }_{0}+{N}_{q,{prod}}left(tright)cdot {left({G}_{q,{non}-{prod}}left(tright)cdot left(1-{kappa }_{{prod},q}right)right)}^{2}cdot left(1-{mu }_{0}right)$$$$-{N}_{q,{prod}}left(tright)cdot {delta }_{0}+{N}_{q,{non}-{prod}}left(tright)cdot {{G}_{q,{non}-{prod}}left(tright)}^{2}cdot {mu }_{0}=0$$Because the mutation terms cancel:$$left({N}_{q,{non}-{prod}}left(tright)+{N}_{q,{prod}}left(tright)cdot {left(1-{kappa }_{{prod},q}right)}^{2}right)cdot left({{G}_{q,{non}-{prod}}left(tright)}^{2}-{delta }_{0}right)=0$$ (26) Substituting in for the growth rate terms (3–7) gives:$$left({N}_{q,{non}-{prod}}(t)+{N}_{q,{prod}}(t)cdot {left(1-{kappa }_{{prod},q}right)}^{2}right)cdot left({left({G}_{0}cdot {R}_{q,{BIOAVAILABLE}}(t)cdot {f}_{{conv}}right)}^{2}-{delta }_{0}right)=0$$ (27) By (16), (4), (6) and the above, total steady state assimilation of the growth limiting environmental substance by species (q) is:$${A}_{{R}_{q}}left(tright)=mathop{sum }limits_{k=1}^{{j}_{{total}}}frac{{G}_{q,k}left(tright)cdot left({N}_{q,k}left(tright)-{rho }_{q,k}left(tright)right)}{{G}_{q,{kPR}}}$$$$kern2.4pc=frac{{N}_{q,{non}-{prod}}left(tright)cdot {left({G}_{q,{non}-{prod}}left(tright)right)}^{2}}{{G}_{0}}+frac{left({N}_{q,{prod}}left(tright)cdot {left(1-{kappa }_{{prod},q}right)}^{2}right)cdot {left({G}_{q,{non}-{prod}}left(tright)right)}^{2}}{{G}_{0}cdot left(1-{kappa }_{{prod}}right)}$$$$=left({N}_{q,{non}-{prod}}left(tright)+{N}_{q,{prod}}left(tright)cdot left(1-{kappa }_{{prod},q}right)right)cdot {{G}_{0}cdot ({R}_{q,{BIOAVAILABLE}}left(tright)cdot {f}_{{conv}})}^{2}$$ (28) By (16–18), production of this substance by the producer allele in the other species (p , ne , q), assuming the various arguments above simultaneously apply to this species, is:$${P}_{{R}_{q}}left(tright)= frac{{G}_{p,{prod}}left(tright)cdot left({N}_{p,{prod}}left(tright)-{rho }_{p,{PRODUCER}}left(tright)right)}{{G}_{p,{prodPR}}}cdot {f}_{{conv},{PROD}}\ = {N}_{p,{prod}}left(tright)cdot left(1-{kappa }_{{prod},p}right) cdot {G}_{0}cdot {({R}_{p,{BIOAVAILABLE}}left(tright)cdot {f}_{{conv}})}^{2}cdot {f}_{{conv},{PROD}}$$ (29) Balance between input and output fluxes of each environmental substance requires ({varphi }_{{R}_{q}}left(tright)+{P}_{{R}_{q}}left(tright)={A}_{{R}_{q}}left(tright)), meaning that by substituting in ({A}_{{R}_{q}}left(tright)) from (28) it is possible to solve for bioavailable substance level, then substitute in the production flux of this substance derived from the producer allele in the other species (p,ne, q):$${R}_{q,{BIO}{AVAILABLE}}left(tright) =frac{1}{{f}!_{{conv}}}sqrt{frac{{varphi }_{{R}_{q}}left(tright)+{P}_{{R}_{q}}left(tright)}{left({N}_{q,{non}-{prod}}left(tright)+{N}_{q,{prod}}left(tright)cdot left(1-{kappa }_{{prod},q}right)right)cdot {G}_{0}}}\ =frac{1}{{f}!_{{conv}}}sqrt{frac{{varphi }_{{R}_{q}}left(tright)+{N}_{p,{prod}}left(tright)cdot left(1-{kappa }_{{prod},p}right)cdot {G}_{0}cdot {({R}_{p,{BIOAVAILABLE}}left(tright)cdot {f}_{{conv}})}^{2}cdot {f}_{{conv},{PROD}}}{left({N}_{q,{non}-{prod}}left(tright)+{N}_{q,{prod}}left(tright)cdot left(1-{kappa }_{{prod},q}right)right)cdot {G}_{0}}}$$ (30) Substituting this into (27) gives a symmetrical condition for steady state genotype frequencies and substance levels across the system:$$ left({N}_{q,{non}-{prod}}left(tright)+{N}_{q,{prod}}left(tright)cdot {left(1-{kappa }_{{prod},q}right)}^{2}right)cdot\ left(frac{{varphi }_{{R}_{q}}left(tright)+{N}_{p,{prod}}left(tright)cdot left(1-{kappa }_{{prod},p}right)cdot {G}_{0}cdot {({R}_{p,{BIOAVAILABLE}}left(tright)cdot {f}_{{conv}})}^{2}cdot {f}_{{conv},{PROD}}}{left({N}_{q,{non}-{prod}}left(tright)+{N}_{q,{prod}}left(tright)cdot left(1-{kappa }_{{prod},q}right)right)cdot {G}_{0}}-{delta }_{0}right)=0$$ (31) This solution illustrates the intuitive ideas that growth and reproduction balance random death at steady state and that the associated producer frequency is lower than that of the non-producer by a factor of the cost. (This factor is of second order because the growth rate is used both directly and (by (10)) in the calculation of starvations). Because our model is a discrete stochastic process, (31) can be viewed as an approximation to a steady state condition, subject to the above assumptions combined with the continuous generation of producers by mutation at a sufficient rate to preclude their extinction. The key point is that over long timescales in the finite populations with which we deal, organism-level selection unavoidably favors the non-producer, with no possibility for multi-level fecundity selection. The producer’s stable presence is thus attributable to the combination of mutation and cycle-level selection. More

Continental, regional, and local spatiotemporal patterns of carbon lossFor Africa’s primary tropical humid forest, carbon losses due to forest disturbances reached 42.2 ± 5.1 MtC yr−1 (mean ± standard deviation, where MtC yr−1 is one million metric tons of carbon loss per year) in 2019 and 53.4 ± 6.5 MtC yr−1 in 2020. Just 9 countries out of the 23 analyzed accounted for 95.0% of total gross losses in 2019 and 94.3% in 2020. These countries contain about 95.7% of all primary tropical humid forests of Africa, with the DRC accounting for 52.8%, Gabon 11.8%, the Republic of the Congo 11.0%, and Cameroon 9.8%. Of these, DRC and Cameroon were responsible for 49.3% and 19.1% of losses in 2019 and 44.7% and 20.6% in 2020. DRC and Cameroon had an annual increase of 15.0% and 36.5% respectively, between 2019 and 2020. From countries with at least 1 MtC emitted in the two years analyzed, Madagascar had the highest annual increase in carbon loss (+153.9%), while Equatorial Guinea is the only country with a decrease in carbon loss (−20.1%). Extending the carbon loss analysis for both past and future will help to better understand these variations and whether the COVID-19 global pandemic had any influence on the general increase between 2019 and 202019. While the absolute numbers for carbon loss estimates should be treated carefully and a sample-based approach should be preferred for an unbiased estimate of absolute numbers20, we focused our analysis on the trends of carbon loss at the continental, country, and local scale (Fig. 1 and Supplementary Fig. 1).Fig. 1: Carbon loss across Africa’s rainforests.We analyzed 23 countries containing primary moist forest. The aboveground carbon stock (green palette) underlies the carbon loss estimations (red palette). Several hotspots can be seen across these regions. The uncertainties of the carbon loss estimations are expressed as standard deviations and shown in Supplementary Fig. 1.Full size imageThe high temporal detail of the analysis revealed various monthly patterns of carbon losses for countries, highly related to local rainfall patterns18 (Fig. 2). Countries like Cameroon, Liberia, Nigeria, Central African Republic (CAR), and Madagascar showed a clear dry-wet seasonal variation in carbon loss per year, while the Republic of the Congo and the DRC, due to their latitudinal extent, exhibited two dry-wet season variations per year with varying intensities (Fig. 2). The seasonal variation can be explained by higher accessibility to forests during the dry months when activities related to smallholder agriculture and logging are more feasible than in the wet season when many roads become inaccessible.Fig. 2: Temporal patterns of carbon loss for the top 10 countries.We show monthly statistics for 2019 and 2020 and the associated uncertainty (black lines). We separate between high (red bars) and low (yellow bars) confidence alerts, the latter showing up for the last 3 months of 2020.Full size imageOne of the highest differences between the months with the most and the least carbon losses was found for Madagascar (72 times more carbon loss in March compared to November 2019). In CAR, the three consecutive months with the highest cumulative carbon loss (January to March 2020) contributed to 75.7% of the total annual loss (between February and April 2020), in Nigeria 73.9% (January to March 2020), Liberia 73.1% (February to April 2020), Madagascar 70.7% (September to November 2020), and Cameroon 62.2% (January to March 2020). Lower percentages were found for countries with mixed seasonality and patterns, like DRC 36.7% (January to March 2020), and the Republic of the Congo 32.8% (January to March 2020) (Fig. 2). For the latter two countries, we expect better-defined peaks of carbon loss at local scales, where climatic conditions are not mixed. The annual cumulative carbon loss (%) per country (Fig. 3) showed that Liberia, Nigeria, CAR, and Cameroon reached between 70-90% of their annual carbon loss in April, while Madagascar reached 60% in October. The DRC, Gabon, Republic of the Congo, Equatorial Guinea, and Ghana have a more gradual monthly increase of cumulative carbon loss with less contrasting seasonality effects. Monthly patterns of carbon losses between the two years analyzed resulted in a correlation coefficient of 0.94 for the CAR, 0.92 for the DRC, 0.91 for Madagascar, 0.90 for Gabon, and 0.83 for Cameroon (Supplementary Fig. 2). For the Republic of the Congo, the two years correlated 0.51. Knowing the peak months of carbon loss for each country and that these patterns are repeatable from one year to another can contribute to better target and prioritize enforcement activities, as well as predicting future patterns and early reporting of annual forest carbon losses.Fig. 3: Annual cumulative carbon loss (%) for both years analyzed, 2019 and 2020.Africa’s total cumulative carbon loss is shown with a black line. The 10 topmost emitting countries out of 23 countries analyzed are shown and represented by distinct colored lines.Full size imageSeveral hotspots of carbon losses can be seen in Fig. 1. The high spatial and temporal details of our analysis are shown in Fig. 4, where several local examples with different drivers of forest disturbances are shown, like logging roads, selective logging, mining, oil palm plantations, urban expansion, and small-holder agriculture. This kind of information, coupled with auxiliary datasets (e.g., legal concessions, protected areas) can identify the legality of forest disturbance21.Fig. 4: Local examples of approx. 10 × 10 km in extent showing different spatiotemporal patterns and drivers of carbon loss.The first column shows the carbon loss, the second column the associated uncertainty, the third column the day-of-the-year when the loss occurred, and the last column shows the monthly distribution of carbon loss and associated uncertainty for each local example. The center coordinates of each location are shown in the third column as latitude and longitude. Exact locations are shown in Supplementary Fig. 3. a Logging roads and selective logging in the Central African Republic, b mining of gold and titanium in the Republic of the Congo, c development of an oil palm plantation in Cameroon, d forest disturbance related to building a new capital city in Equatorial Guinea, and e small-scale agriculture expansion at the edge of the forest in the DRC.Full size imageImplications of rapid monitoring of local carbon lossNear-real time alerts combined with biomass maps result in spatially explicit forest carbon loss, unlike global tabular statistics of national data22,23. We provide new insights into the spatiotemporal dynamics of carbon loss with consistent assessment of accuracy that could enable transparency and completeness for countries reporting on their REDD + progress to the UNFCCC24. We provide monthly carbon loss estimates that could play a key role in local, national, and international forest initiatives for global carbon policy goals25. Such a system can be implemented with minimal costs and is based on open-source datasets and Google Earth Engine cloud computing platform26, thus enabling cost-effective national monitoring of forest carbon loss7. Providing rapid reporting on the location, time, and amount of carbon lost across Africa’s primary humid forest will help undertake immediate action to protect and conserve carbon-rich threatened forests. Furthermore, countries will be able to predict and estimate their annual carbon loss before a reporting period ends, thus having the opportunity to adjust their practices to meet their country-specific commitments for climate change mitigation initiatives.Limitations and future improvementsWe used the RADD alerts (Radar for Detecting Deforestation)18 with a minimum mapping unit (MMU) of 0.2 ha as accuracy estimates were available for this MMU. Events smaller than 0.2 ha would add to the total carbon loss but are by nature associated with higher uncertainties18. The implications of the RADD alerts using a global humid tropical forest product as a forest baseline for 201816,27,28 are twofold. First, the global nature of this product might result in inconsistencies at the local level18. Second, because the forest cover loss information used to generate the forest baseline is based on optical Landsat data, persistent cloud cover in the second half of 2018 in some areas led to missed reporting of forest disturbances, thus being detected at the beginning of 2019 by the RADD alerts. This possible overestimation of carbon loss at the start of 2019 is not an issue for a near-real-time alerting system since later months are not affected. Furthermore, the alerts do not distinguish between human-induced disturbances and natural forest disturbances18. When a new forest disturbance alert is detected, it will be confirmed or rejected within 90 days by subsequent Sentinel-1 images18. That is why our carbon loss reporting separates between high and low confidence alerts for the last three months of 2020, which is common for most forest disturbance alerting products18,29. We separated all the alerts into core and boundary pixels. Core alerts represent complete tree cover removal and we assumed complete carbon loss within a pixel. For boundary alerts, we assumed a 50% carbon loss since these mainly represent forest disturbances with partial tree cover removal. Detecting and quantifying the level of degradation remains challenging and future developments will minimize this uncertainty by providing variable percentages of degraded forest30. The timeliness and spatial details of future forest disturbance alerting products will improve with the availability of open access long-wavelength radar data from near-future satellite missions (e.g., NISAR L-band SAR in 202331), by using a combination of optical and radar forest disturbance alert products, and integration with high-resolution satellite products.We relied on an aboveground biomass baseline map from 201832, prior to RADD alerts starting from 2019. Biomass estimation for the tropical moist forests is based on ALOS-2 PALSAR-2 L-band satellite and its usage needs to account for the local biases, especially underestimating AGB values higher than 250 Mg ha−1 (ref. 32). Although we reduced this underestimation by adjusting the AGB map based on ground field data, more research is needed on providing up-to-date high-resolution aboveground carbon estimates33 that could further increase the accuracy of local carbon loss estimation. Radar-based estimation of forest carbon stocks is challenging over mountainous terrain and is less accurate in complex canopies3 and future integration of radar and optical satellite data will provide more robust estimates33. Nevertheless, new spaceborne missions (e.g., GEDI34, BIOMASS35) will provide an unprecedented amount of forest structure samples that will improve the algorithms and thus the final accuracy of aboveground biomass estimates.We focused on exploring and analyzing local carbon losses and showing high temporal and spatial patterns of carbon losses. We showed the country statistics to emphasize the temporal dynamics of carbon losses and compare the temporal profiles across our study region. Our approach was not to provide stratified area estimations36 associated with forest disturbances but we used this concept in the sense that we had a stratified sample of higher quality reference data18 to estimate the omission and commission errors and consider those in our uncertainty estimation on the pixel level. The analysis showed that omission and commission errors are small and rather balanced, and thus do not result in a major area bias for the forest disturbances. The uncertainties of the aboveground biomass product32 were adjusted for known regional biases using regional forest biomass plot data sources. With this approach, the original aboveground biomass map bias was partly corrected using a model-based approach deemed to be an alternative to a sample-based approach whenever country data are unavailable37. Our uncertainty analysis and error reduction showed that we expect only minor bias in the forest disturbance and the biomass data and the remaining uncertainties are propagated in our pixel-based uncertainty layer. More