Philippa Kaur

The generation of this map of the world’s mountains consisted of five steps (Fig. 1): (i) the identification and hierarchisation of named mountain ranges and the recording of range-specific information; (ii) the manual digitization of the ranges’ general shape; (iii) the definition of mountainous terrain (and the inventory’s outer borders) using a DEM-based algorithm; (iv) the automatic refinement of the digitized and named ranges’ inner borders; and (v) the preparation of the final layers. The resulting products consist of a refined mountain definition (GMBA Definition v2.0), two versions of the inventory (GMBA Inventory v2.0_standard & GMBA Inventory v2.0_broad), and a set of tools to work with the inventories.Step i: Identification and hierarchisation of mountain rangesIn a first step, we identified mountain ranges worldwide. To do so we adopted the mountain ranges identified in the GMBA Inventory v1.410,14 and searched existing resources in any languages for other named ranges not yet included. The ranges added could either be adjacent to, included in (child range or subrange) or including (parent range or mountain system) mountain ranges of the GMBA Inventory v1.4. The resources used for our searches included world atlases (e.g. The Times Comprehensive Atlas of the World19, Knaurs grosser Weltatlas20, Pergamon World Atlas21); topographic maps (e.g. http://legacy.lib.utexas.edu/maps/imw/, http://legacy.lib.utexas.edu/maps/onc/, https://maps.lib.utexas.edu/maps/tpc/, www.topomap.co.nz, https://norgeskart.no, www.ign.es/iberpix/visor/); encyclopaedias (www.wikipedia.org; www.britannica.com); online gazetteers and reference sites (e.g. www.wikidata.org, www.geonames.org (GeoNames), www.mindat.org); mountain classification systems (e.g. the International Standardized Mountain Subdivision of the Alps or SOIUSA for the Alps22, Alpenvereinseinteilung der Ostalpen23, Classification of the Himalaya24, www.peakbagger.com/rangindx.aspx (PEMRACS), www.carpathian-research-network.eu/ogulist, http://www.sopsr.sk/symfony-bioregio/lkpcarporog, www.dinarskogorje.com, https://bivouac.com/, https://climbnz.org.nz/); and national or regional landscape, geomorphological, or physiographic maps and publications4,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42. The full list of the consulted sources and references is available on GitHub at https://www.github.com/GMBA-biodiversity/Inventory (GMBA Mountain Inventory v2.0 References.pdf).All identified mountain ranges were recorded in a Microsoft Access relational database (“Mountain database”, see below) and given a name, a unique 5-digit identifier (GMBA_V2_ID), and the corresponding Wikidata unique resource identifier (URI), when available. This URI gives access to a range’s name as well as to its Wikipedia page URL in all available languages and lists other identifiers for given mountain ranges in a variety of other repositories such as GeoNames or PEMRACS. The primary mountain range names were based on the resources used for range identification and were preferably recorded in English. Names used nationally, locally, as well as/or by indigenous people and local communities were extracted from Wikidata and recorded in a separate attribute field.In the process of cataloguing, we attributed a parent range to each of the mapped mountain ranges. Information about parent ranges is included in PEMRACS, often also in Wikidata as a property that can be extracted though a SPARQL query, in the corresponding Wikipedia pages description, and in regional hierarchical mountain classifications that exist for the European Alps (SOIUSA), the Carpathians, and the Dinaric Alps. When no such information was available, we relied on other sources of information that we found either using a general web search (leading to specific papers, reports, or web pages on mountain ranges) or by consulting (online) topographical maps and atlases at different scales. The information about parent ranges was used to construct a hierarchy of up to 10 levels using a recursive SQL query (see Step v). The result of this step was a relational database with a hierarchy of mountain systems and (sub-) ranges (Fig. 1, “Mountain database”).Step ii: Digitization of the mountain rangesIn a second step, we digitized all identified ‘childless’ mountain ranges (i.e. smallest mapping units, called ‘Basic’ as opposed to ‘Aggregated’ in the database) in one vector GIS layer. To do so, we used the Google Maps Terrain layers (Google, n.d.) as background and the WHYMAP named rivers layer42 as spatial reference since descriptions of mountain range areal extension is often given with reference to major rivers. The digitization, which was done in QGIS43 using the WGS 84 / Pseudo-Mercator (EPSG 3857) coordinate reference system, consisted in the drawing of shapes (polygons) that roughly followed the core area of each mountain range. In general, the approximate shape and extent of the mountain ranges we digitized could be distinguished based on the terrain structure as represented by the shaded relief background that corresponded to the placement and orientation of the range’s name label on a topographical map, atlas or other resource. As the exact placement and orientation of mountain range labels in each specific source can be influenced by cartographic considerations (e.g. avoiding overlaps with other features), the final approximation of the mountain range was obtained by consulting a variety of sources for each mountain range. Occasionally, the mountain terrain’s geomorphological characteristics strongly hampered the accuracy of our visual identification of mountain subranges within larger systems. This was particularly the case in old, eroded massifs such as the Brazilian Highlands or the highlands of Madagascar, where individual mountain ranges are not separated by deep well-defined valleys and have a very complex topography. In these cases, we referred to available topographical descriptions of range extent and to the river layer (see above). Other complex regions included Borneo and the Angolan Highlands, whereas subranges in mountain systems such as the European Alps, the Himalayas, and the North American Cordillera were comparatively easy to map. Moreover, the density of currently available mountain toponymical information varied quite strongly between regions. Accordingly, regional variation in the size of the smallest mountain range map units can be considerable. The result of this step was a (manually) digitized vector layer of named mountain ranges shapes (Fig. 1, “Manual mountain shapes”).Step iii: Definition of mountainous terrainIn a third step, we defined mountainous terrain (GMBA Definition v2.0). To distinguish mountainous from non-mountainous terrain, we developed a simple algorithm which we implemented in ArcMap 10.7.144. This algorithm is based on ruggedness (defined as highest minus lowest elevation in meter) within eight circular neighbourhood analysis windows (NAWs) of different sizes (from 1 pixel (≈ 250 m) to 20 (≈ 5 km) around each point, Fig. 2c) combined with empirically derived thresholds for each NAW (Fig. 2). The decision to use multiple NAW sizes was made because calculating ruggedness based on only a small or a large NAW comes at the risk of identifying the many local irregularities typically occurring in flat or rolling terrain as mountainous or of including extensive flat ‘skirts’ through the smoothing and generalization of large NAWs3. Accordingly, our approach ensures that any point in the landscape classified as mountainous showed some level of ruggedness not only at one but across scales. This also resulted in a smooth and homogeneous delineation of mountainous terrain, very suitable for our mapping purpose.Fig. 2Elevation range thresholds for the eight neighbourhood analysis windows (NAW) and their contribution to calculations of the GMBA Definition v2.0. (a) distribution of elevation range values (ruggedness) for NAWs (numbered I to VIII) in mountain regions as defined by the geometric intersection of K1, K2 and K3. (b): plot of the minimum elevation range versus the area of the NAW (n = 920). (c) NAWs and their corresponding threshold values. (d) percent overlap between GMBA Definition v2.0 (intersection of eight NAW-threshold pairs) and area defined by each individual NAW-threshold pair. (e) percent eliminated by each NAW-threshold pair (I to VIII) from the mountain area defined by the other 7 NAW-threshold combinations. Highlighted bars in the two graphs represent the combination of three NAW-threshold pairs that results in the highest overlap with the GMBA Definition v2.0.Full size imageWe used the median value of the 7.5 arc second GMTED2010 DEM45 as our source map. To reduce the latitudinal distortion of the raster, and thus the shape and area of the NAWs, we divided the global DEM into three raster layers corresponding to three latitudinal zones (84° N to 30° N, 30° N to 30° S and 30° S to 56° S) excluding ice-covered Antarctica and projected the two high latitude zones to Lambert Azimuthal Equal Area and the equatorial zone to WGS 1984 Cylindrical Equal Area. We used these reprojected DEM layers to produce eight ruggedness layers, each using one of the eight NAWs.To determine the threshold values of our algorithm, we selected 1000 random points within the area defined by the geometric intersection (Fig. 1b) of the three commonly applied mountain definitions, i.e. the definitions by UNEP-WCMC46, GMBA15, and USGS3. These layers (referred to as K1, K2, and K3, respectively by Sayre and co-authors12) were obtained from the Global Mountain Explorer47. We eliminated 80 clearly misclassified points (i.e., points that fell within lakes, oceans, or clearly flat areas according to the shaded relief map we used as a background) and used the remaining 920 to sample the eight ruggedness layers. For each of the 8 layers, we retained the lowest of the 920 ruggedness values as the threshold for the layer’s specific NAW (Fig. 2c). The eight threshold values were then used to reclassify each of the eight layers by attributing the value 1 to all cells with a ruggedness value higher than or equal to the corresponding threshold and the value 0 to all other cells. Finally, we performed a geometric intersection (see Fig. 1b) of the eight reclassified layers to derive the new mountain definition.After these calculations, we reprojected the three raster layers to WGS84 and combined them through mosaic to new raster. To eliminate isolated cells and jagged borders, we then generalized the resulting raster map by passing a majority filter (3 × 3 pixels, majority threshold) three times. This layer corresponds to the GMBA Definition v2.0.The resulting mountain definition (GMBA Definition v2.0) distinguishes itself from previous ones because of the empirically derived thresholds method used to develop it and the use of eight NAWs. In line with the previous GMBA definition, it relies entirely on the ruggedness values within NAWs. The GMBA Definition v2.0 was used to determine the outer delineation of this inventory’s mountainous terrain. As expected, it includes neither the wide ‘skirts’ of flat or undulating land around mountain ranges nor the topographical irregularities that are both typically included when other approaches are applied. It also successfully excludes extensive areas of rolling non-mountainous terrain such as the 52,000 km2 Badain Jaran Desert sand dunes in China. However, this mountain definition is conservative and only includes the highest, most rugged cores of low mountain systems, as for example in the Central Uplands of Germany, and therefore excludes some lower hill areas still considered by some as mountains.As a further step towards generalization, we considered that small ( More

Tagging deployments and study subjectsTable 1 contains an overview of the fields in the metadata file (JWS_metadata.xlsx) providing extensive background details on each of the 79 tag deployments and 63 study subjects. The data in this file give essential contextual information needed to understand the methodological, environmental, and demographic factors surrounding the deployments, which are critical for further examination and hypothesis testing of the sensor data. These metadata fall into several specific categories, but are not limited to, (i) information on the deployed electronic devices (platform, model, Platform Transmitter Terminal identifications), (ii) sharks (unique identifying numbers, sex, length), (iii) capture event (date, location, duration, methodology, interaction type), and (iv) the reporting period (duration, linear surface travel distance).Table 1 Metadata descriptions of the sharks, tagging operations, and deployments for all tags included in the database.Full size tableFigure 1 illustrates a typical C. carcharias tagging operation. This involves a contracted commercial fishing vessel with purpose-built gears to capture sharks (Fig. 1a) and a research crew to handle animals, monitor health (Fig. 1b) and attach electronic tags (Fig. 1c). More details on the tagging program and its methodologies are provided elsewhere14,19,20. Figure. 2 provides summaries of the deployment schedule, geographic locations, devices, and capture operations. Of note, 39.7% (25/64) of all tagging operations involved collaborations with commercial fishery operators (Fig. 2f–h), whose engagement was temporarily impacted (Fig. 2a) during the scientific review process when the population was under consideration for US Endangered Species Act listing. Figure 3 displays the demographic focus on small juvenile C. carcharias, with modest deployment durations and travel distances.Fig. 1Depiction of a typical research operation for capturing and tagging juvenile White Sharks in the Southern California Bight. (a) Aquarium research vessel (RV Lucile) with crew approaching a contracted purse seine vessel containing a captured juvenile white shark. (b) Research crew on the RV Lucile leading the shark into a sling, where it is subsequently transferred to the vessel’s deck for tagging. (c) Successfully applied PAT and acoustic tags each positioned lateral of the dorsal fin, anchored via leaders, and affixed with titanium darts (yellow arrows). All images taken by Steve McNicholas (Great White Shark 3D) for the Monterey Bay Aquarium and used with permission.Full size imageFig. 2Metadata summaries of the field program that deployed biologging tags on juvenile white sharks in the southern California Current. (a) Deployment schedule for 72 electronic tags released on 64 White Sharks from 2001–2020 (b) Tagging activity peaked in the late summer months when the population is most locally abundant. Field operations decreased from 2011–2013 when the population was being considered for listing under the U.S. Endangered Species Act (ESA). (c) Deployments focused on opportunities in the Southern California Bight coastline and included deployments in the nursery area of Bahía Sebastian Vizcaíno, Mexico and releases after exhibition at the Monterey Bay Aquarium. (d) Researchers released a variety of pop-up archival transmitting (PAT, 58 sharks), acoustic (21 sharks), and smart position and temperature (SPOT, 20 sharks) tags. This manuscript only reports the geolocation, temperature and depth data from the PAT and SPOT platforms. (e) Half (35 of 64, 54.7%) of all sharks received multiple tags, primarily to compare their relative performance. (f) Most tags (38 of 64, 60.3%) were deployed during focused scientific research operations. (g) The remainder were joint operations resulting from opportunistic bycatch in commercial fisheries using various gears and (h) Targeting various species. “Jab” gear refers to research operations that uses pole extensions to apply tags to sharks without capturing and handling.Full size imageFig. 3Demographic and deployment summaries from the juvenile white shark tagging program. (a) Total body length (TL) histogram indicates that most individuals tagged were either neonates ( More

FAO & UNEP. The state of the world’s forests 2020: Forests, biodiversity and people (2020).Brinck, K. et al. High resolution analysis of tropical forest fragmentation and its impact on the global carbon cycle. Nature Communications 8, 1–6 (2017).ADS 

Google Scholar 
Mitchard, E. T. The tropical forest carbon cycle and climate change. Nature 559, 527–534 (2018).ADS 
CAS 
PubMed 

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

Google Scholar 
Tsujino, R., Yumoto, T., Kitamura, S., Djamaluddin, I. & Darnaedi, D. History of forest loss and degradation in Indonesia. Land use policy 57, 335–347 (2016).
Google Scholar 
Holl, K. D. & Brancalion, P. H. Tree planting is not a simple solution. Science 368, 580–581 (2020).ADS 
CAS 
PubMed 

Google Scholar 
Zhang, Y. et al. Multiple afforestation programs accelerate the greenness in the ‘Three North’ region of China from 1982 to 2013. Ecological Indicators 61, 404–412 (2016).
Google Scholar 
Soliño, M., Oviedo, J. L. & Caparrós, A. Are forest landowners ready for woody energy crops? Preferences for afforestation programs in Southern Spain. Energy Economics 73, 239–247 (2018).
Google Scholar 
Paquette, A. & Messier, C. The role of plantations in managing the world’s forests in the Anthropocene. Frontiers in Ecology and the Environment 8, 27–34 (2010).
Google Scholar 
Zulkefli, F., Syahlan, S. & Aziz, M. F. A. Negatives Impact Faced by Oil Palm Estate Management in managing Foreign Workers: A Case Study. International Journal of Academic Research in Business and Social Sciences 8 (2018).Fitzherbert, E. B. et al. How will oil palm expansion affect biodiversity? Trends in ecology & evolution 23, 538–545 (2008).
Google Scholar 
Vijay, V., Pimm, S. L., Jenkins, C. N. & Smith, S. J. The impacts of oil palm on recent deforestation and biodiversity loss. PloS one 11, e0159668 (2016).PubMed 
PubMed Central 

Google Scholar 
Koh, L. P., Miettinen, J., Liew, S. C. & Ghazoul, J. Remotely sensed evidence of tropical peatland conversion to oil palm. Proceedings of the National Academy of Sciences 108, 5127–5132 (2011).ADS 
CAS 

Google Scholar 
Guillaume, T. et al. Carbon costs and benefits of Indonesian rainforest conversion to plantations. Nature communications 9, 1–11 (2018).CAS 

Google Scholar 
Lucas-Borja, M. E., Hedo, J., Cerdá, A., Candel-Pérez, D. & Viñegla, B. Unravelling the importance of forest age stand and forest structure driving microbiological soil properties, enzymatic activities and soil nutrients content in Mediterranean Spanish black pine (Pinus nigra Ar. ssp. salzmannii) Forest. Science of the Total Environment 562, 145–154 (2016).ADS 
CAS 

Google Scholar 
Besnard, S. et al. Quantifying the effect of forest age in annual net forest carbon balance. Environmental Research Letters 13, 124018 (2018).ADS 

Google Scholar 
Dzikiti, S. et al. Estimating the water requirements of high yielding and young apple orchards in the winter rainfall areas of South Africa using a dual source evapotranspiration model. Agricultural water management 208, 152–162 (2018).
Google Scholar 
Zhang, Y., Yao, Y., Wang, X., Liu, Y. & Piao, S. Mapping spatial distribution of forest age in China. Earth and Space Science 4, 108–116 (2017).ADS 

Google Scholar 
Chen, B., Jin, Y. & Brown, P. Automatic mapping of planting year for tree crops with Landsat satellite time series stacks. ISPRS Journal of Photogrammetry and Remote Sensing 151, 176–188 (2019).ADS 

Google Scholar 
Danylo, O. et al. A map of the extent and year of detection of oil palm plantations in Indonesia, Malaysia and Thailand. Scientific data 8, 1–8 (2021).
Google Scholar 
O’Brien, S. T., Hubbell, S. P., Spiro, P., Condit, R. & Foster, R. B. Diameter, height, crown, and age relationship in eight neotropical tree species. Ecology 76, 1926–1939 (1995).
Google Scholar 
Fichtler, E., Clark, D. A. & Worbes, M. Age and long-term growth of trees in an old-growth tropical rain forest, based on analyses of tree rings and 14C1. Biotropica 35, 306–317 (2003).
Google Scholar 
Zhang, C. et al. Mapping forest stand age in China using remotely sensed forest height and observation data. Journal of Geophysical Research: Biogeosciences 119, 1163–1179 (2014).ADS 

Google Scholar 
Wang, B., Li, M., Fan, W., Yu, Y. & Chen, J. M. Relationship between net primary productivity and forest stand age under different site conditions and its implications for regional carbon cycle study. Forests 9, 5 (2018).
Google Scholar 
Wang, S. et al. Relationships between net primary productivity and stand age for several forest types and their influence on China’s carbon balance. Journal of environmental management 92, 1651–1662 (2011).PubMed 

Google Scholar 
Gupta, N., Kukal, S., Bawa, S. & Dhaliwal, G. Soil organic carbon and aggregation under poplar based agroforestry system in relation to tree age and soil type. Agroforestry Systems 76, 27–35 (2009).
Google Scholar 
Huang, C. et al. An automated approach for reconstructing recent forest disturbance history using dense Landsat time series stacks. Remote Sensing of Environment 114, 183–198 (2010).ADS 

Google Scholar 
Thomas, N. E. et al. Validation of North American forest disturbance dynamics derived from Landsat time series stacks. Remote Sensing of Environment 115, 19–32 (2011).ADS 
MathSciNet 

Google Scholar 
Ye, S., Rogan, J., Zhu, Z. & Eastman, J. R. A near-real-time approach for monitoring forest disturbance using Landsat time series: Stochastic continuous change detection. Remote Sensing of Environment 252, 112167 (2021).ADS 

Google Scholar 
Verbesselt, J., Hyndman, R., Newnham, G. & Culvenor, D. Detecting trend and seasonal changes in satellite image time series. Remote sensing of Environment 114, 106–115 (2010).ADS 

Google Scholar 
Kennedy, R. E., Yang, Z. & Cohen, W. B. Detecting trends in forest disturbance and recovery using yearly Landsat time series: 1. LandTrendr—Temporal segmentation algorithms. Remote Sensing of Environment 114, 2897–2910 (2010).ADS 

Google Scholar 
Cohen, W. B., Yang, Z., Healey, S. P., Kennedy, R. E. & Gorelick, N. A LandTrendr multispectral ensemble for forest disturbance detection. Remote sensing of environment 205, 131–140 (2018).ADS 

Google Scholar 
Vogeler, J. C., Braaten, J. D., Slesak, R. A. & Falkowski, M. J. Extracting the full value of the Landsat archive: Inter-sensor harmonization for the mapping of Minnesota forest canopy cover (1973–2015). Remote sensing of environment 209, 363–374 (2018).ADS 

Google Scholar 
de Jong, S. M. et al. Mapping mangrove dynamics and colonization patterns at the Suriname coast using historic satellite data and the LandTrendr algorithm. International Journal of Applied Earth Observation and Geoinformation 97, 102293 (2021).
Google Scholar 
Harris, N., Goldman, E. D. & Gibbes, S. Spatial database of planted trees (SDPT VERSION 1.0). Technical Note. (2019).Descals, A. et al. High-resolution global map of smallholder and industrial closed-canopy oil palm plantations. Earth System Science Data 13, 1211–1231 (2021).ADS 

Google Scholar 
Li, C. et al. The first all-season sample set for mapping global land cover with landsat-8 data. Science Bulletin 62, 508–515 (2017).ADS 

Google Scholar 
Masek, J. G. et al. A Landsat surface reflectance dataset for North America, 1990–2000. IEEE Geoscience and Remote Sensing Letters 3, 68–72 (2006).ADS 

Google Scholar 
Vermote, E., Justice, C., Claverie, M. & Franch, B. Preliminary analysis of the performance of the Landsat 8/OLI land surface reflectance product. Remote Sensing of Environment 185, 46–56 (2016).ADS 
PubMed 

Google Scholar 
Foga, S. et al. Cloud detection algorithm comparison and validation for operational Landsat data products. Remote sensing of environment 194, 379–390 (2017).ADS 

Google Scholar 
Goulden, M. L. & Bales, R. C. California forest die-off linked to multi-year deep soil drying in 2012–2015 drought. Nature Geoscience 12, 632–637 (2019).CAS 

Google Scholar 
He, T. et al. Evaluating land surface albedo estimation from Landsat MSS, TM, ETM+, and OLI data based on the unified direct estimation approach. Remote Sensing of Environment 204, 181–196 (2018).ADS 

Google Scholar 
Flood, N. Continuity of reflectance data between Landsat-7 ETM+ and Landsat-8 OLI, for both top-of-atmosphere and surface reflectance: a study in the Australian landscape. Remote Sensing 6, 7952–7970 (2014).ADS 

Google Scholar 
Roy, D. P. et al. Characterization of Landsat-7 to Landsat-8 reflective wavelength and normalized difference vegetation index continuity. Remote sensing of Environment 185, 57–70 (2016).ADS 
PubMed 

Google Scholar 
Key, C. & Benson, N. Landscape assessment: remote sensing of severity, the normalized burn ratio and ground measure of severity, the composite burn index. FIREMON: Fire effects monitoring and inventory system Ogden, Utah: USDA Forest Service, Rocky Mountain Res. Station (2005).Guo, J. & Gong, P. The potential of spectral indices in detecting various stages of afforestation over the Loess Plateau Region of China. Remote Sensing 10, 1492 (2018).ADS 

Google Scholar 
Kennedy, R. E. et al. Implementation of the LandTrendr algorithm on google earth engine. Remote Sensing 10, 691 (2018).ADS 

Google Scholar 
Yu, L. et al. A multi-resolution global land cover dataset through multisource data aggregation. Science China Earth Sciences 57, 2317–2329 (2014).ADS 

Google Scholar 
Du, Z. et al. A global map of planting years of plantations. figshare https://doi.org/10.6084/m9.figshare.19070084.v1 (2022).Gong, P. et al. Finer resolution observation and monitoring of global land cover: First mapping results with landsat tm and etm+ data. International Journal of Remote Sensing 34, 2607–2654 (2013).ADS 

Google Scholar 
Huang, H. et al. The migration of training samples towards dynamic global land cover mapping. ISPRS Journal of Photogrammetry and Remote Sensing 161, 27–36 (2020).ADS 

Google Scholar  More

Flores, C. O. et al. Proc. Natl Acad. Sci. USA 108, 288–297 (2011).Article 

Google Scholar 
Carlson, M. C. G. et al. Nat. Microbiol. https://doi.org/10.1038/s41564-022-01088-x (2022).Article 

Google Scholar 
Flombaum, P. et al. Proc. Natl Acad. Sci. USA 110, 9824–9829 (2013).CAS 
Article 

Google Scholar 
Coleman, M. L. & Chisholm, S. W. Trends Microbiol. 15, 398–407 (2007).CAS 
Article 

Google Scholar 
Johnson, Z. I. et al. Science 311, 1737–1740 (2006).CAS 
Article 

Google Scholar 
Martiny, A. C. et al. PLoS ONE 11, e0168291 (2016).Article 

Google Scholar 
Wilhelm, S. W. & Suttle, C. A. Bioscience 49, 781–788 (1999).Article 

Google Scholar 
Follett, C. L. et al. Proc. Natl Acad. Sci. USA 119, e2110993118 (2022).CAS 
Article 

Google Scholar 
Mojica, K. D. A. et al. ISME J. 10, 500–513 (2016).CAS 
Article 

Google Scholar 
Mruwat, N. et al. ISME J. 15, 41–54 (2021).CAS 
Article 

Google Scholar  More

NEWS
01 April 2022

Tropical forests have big climate benefits beyond carbon storage

Study finds that trees cool the planet by one-third of a degree through biophysical mechanisms such as humidifying the air.

Freda Kreier

Freda Kreier

View author publications

You can also search for this author in PubMed
 Google Scholar

Twitter

Facebook

Email

Tropical forests create cloud cover that reflects sunlight and cools the air.Credit: Thomas Marent/Minden Pictures

Tropical forests have a crucial role in cooling Earth’s surface by extracting carbon dioxide from the air. But only two-thirds of their cooling power comes from their ability to suck in CO2 and store it, according to a study1. The other one-third comes from their ability to create clouds, humidify the air and release cooling chemicals.
How much can forests fight climate change?
This is a larger contribution than expected for these ‘biophysical effects’ says Bronson Griscom, a forest climate scientist at the non-profit environmental organization Conservation International, headquartered in Arlington, Virginia. “For a while now, we’ve assumed that carbon dioxide alone is telling us essentially all we need to know about forest–climate interactions,” he says. But this study confirms that tropical forests have other significant ways of plugging into the climate system, he says.The analysis, published in Frontiers in Forests and Global Change on 24 March1, could enable scientists to improve their climate models, while helping governments to devise better conservation and climate strategies.The findings underscore growing concerns about rampant deforestation across the tropics. Scientists warn that one-third of the world’s tropical forests have been mown down in the past few centuries, and another one-third has been degraded by logging and development. This, when combined with climate change, could transform vast swathes of forest into grasslands2.“This study gives us even more reasons why tropical deforestation is bad for the climate,” says Nancy Harris, forest-research director at the World Resources Institute in Washington DC.More than a carbon spongeForests are major players in the global carbon cycle because they soak up CO2 from the atmosphere as they grow. Tropical forests, in particular, store around one-quarter of all terrestrial carbon on the planet, making them “centrepieces for climate policy” in their home countries, Griscom says.
Tropical forests may be carbon sources, not sinks
“There’s clear evidence that the tropics are producing excellent climate benefits for the entire planet,” says Deborah Lawrence, an environmental scientist at the University of Virginia in Charlottesville and a co-author of the latest study. She and her colleagues analysed the cooling capacity of forests around the globe, in particular considering biophysical effects alongside carbon storage. Tropical forests, they found, can cool Earth by a whole 1 °C — and biophysical effects contribute significantly.Although scientists knew about these effects, they hadn’t understood to what extent the various factors counter global warming.Trees in the tropics provide shade, but they also act as giant humidifiers by pulling water from the ground and emitting it from their leaves, which helps to cool the surrounding area in a way similar to sweating, Griscom says.“If you go into a forest, it immediately is a considerably cooler environment,” he says.This transpiration, in turn, creates the right conditions for clouds, which like snow and ice in the Arctic, can reflect sunlight higher into the atmosphere and further cool the surroundings. Trees also release organic compounds — for example, pine-scented terpenes — that react with other chemicals in the atmosphere to sometimes create a net cooling effect.Locally coolTo quantify these effects, Lawrence and her colleagues compared how the various effects of forests around the world feed into the climate system, breaking down their contributions in bands of ten degrees of latitude. When they considered only the biophysical effects, the researchers found that the world’s forests collectively cool the surface of the planet by around 0.5 °C.
When will the Amazon hit a tipping point?
Tropical forests are responsible for most of that cooling. But this band of trees across Latin America, Central Africa and southeast Asia is under increasing pressure from climate change and deforestation. Both of these human-caused impacts can lead rainforests to dry out, says Christopher Boulton, a geographer at the University of Exeter, UK. Last month, he and his colleagues published a review2 of nearly 30 years’ worth of satellite images of the Amazon, the largest rainforest in the world. By measuring the biomass of the vegetation in the images, the team discovered that three-quarters of the Amazon is losing resilience — the ability to recover from an extreme weather event such as a drought.Threats to tropical rainforests are dangerous not only for the global climate, but also for communities that neighbour the forests, Lawrence says. She and her colleagues found that the cooling caused by biophysical effects was especially significant locally. Having a rainforest nearby can help to protect an area’s agriculture and cities from heatwaves, Lawrence says. “Every tenth of a degree matters in limiting extreme weather. And where you have forests, the extremes are minimized.”Governments across the tropics have struggled to conserve their forests despite more than two decades of global campaigns to halt deforestation, promote sustainable development and protect the climate. Lawrence says that her team’s findings make it clear that protecting forests is a matter of self-interest, and has immediate benefits for local communities.

doi: https://doi.org/10.1038/d41586-022-00934-6

ReferencesLawrence, D., Coe, M., Walker, W., Verchot, L. & Vandecar, K. Front. For. Glob. Change https://doi.org/10.3389/ffgc.2022.756115 (2022).Article 

Google Scholar 
Boulton, C. A., Lenton, T. M. & Boers, N. Nature Clim. Change 12, 271–278 (2022).Article 

Google Scholar 
Download references

Related Articles

How much can forests fight climate change?

When will the Amazon hit a tipping point?

Tropical forests may be carbon sources, not sinks

Illegal mining in the Amazon hits record high amid Indigenous protests

Subjects

Climate sciences

Climate change

Conservation biology

Latest on:

Climate sciences

Funding battles stymie ambitious plan to protect global biodiversity
News 31 MAR 22

Trends in Europe storm surge extremes match the rate of sea-level rise
Article 30 MAR 22

From the archive: fishy business in 1972 and 1922
News & Views 29 MAR 22

Climate change

Funding battles stymie ambitious plan to protect global biodiversity
News 31 MAR 22

Trends in Europe storm surge extremes match the rate of sea-level rise
Article 30 MAR 22

The race to upcycle CO2 into fuels, concrete and more
News Feature 29 MAR 22

Jobs

Postdoctoral Fellow in Electrochemical CO2 reduction

The University of British Columbia (UBC)
Kelowna, Canada

Staff Member in Project Coordination (for 19 h/week)

Jülich Research Centre (FZJ)
Erlangen-Nürnberg, Germany

Student assistant IT administration (m/f/d)

Alfred Wegener Institute – Helmholtz Centre for Polar and Marine Research (AWI)
Potsdam, Germany

Physicist (postdoctoral researcher) (all genders) in the field of Theoretical Astrophysics

Helmholtz Centre for Heavy Ion Research GmbH (GSI)
Darmstadt, Germany More

Kwiatkowski, L. et al. Twenty-first century ocean warming, acidification, deoxygenation, and upper-ocean nutrient and primary production decline from CMIP6 model projections. Biogeosciences 17, 3439–3470 (2020).ADS 
CAS 

Google Scholar 
Intergovernmental Panel on Climate Change. Climate Change 2013: 5th Assessment Report of the Intergovernmental Panel on Climate Change (Cambridge University Press, 2013).
Google Scholar 
Torres, O., Kwiatkowski, L., Sutton, A. J., Dorey, N. & Orr, J. C. Characterizing mean and extreme diurnal variability of ocean CO2 system variables across marine environments. Geophys. Res. Lett. 48, 2 (2021).
Google Scholar 
Dorey, N., Lançon, P., Thorndyke, M. & Dupont, S. Assessing physiological tipping point of sea urchin larvae exposed to a broad range of pH. Glob. Change Biol. 19, 3355–3367 (2013).
Google Scholar 
Hauri, C. et al. Spatiotemporal variability and long-term trends of ocean acidification in the California current system. Biogeosci. Discuss. 9, 10371–10428 (2012).ADS 

Google Scholar 
Dupont, S. & Pörtner, H.-O. A snapshot into ocean acidification research. Mar. Biol. 160, 1765–1771 (2013).CAS 

Google Scholar 
Dupont, S. & Thorndyke, M. Chapter: Direct impacts of near-future ocean acidification on sea urchins. in Climate Change Perspective from the Atlantic: Past, Present and Future (eds. Fernández-Palacios, J. et al.) 461–485 (2013).Byrne, M. & Hernández, J. C. Chapter 16: Sea urchins in a high CO2 world: Impacts of climate warming and ocean acidification across life history stages. in Developments in Aquaculture and Fisheries Science vol. 43 281–297 (Elsevier, 2020).Kroeker, K. J., Kordas, R. L., Crim, R. N. & Singh, G. G. Meta-analysis reveals negative yet variable effects of ocean acidification on marine organisms. Ecol. Lett. 13, 1419–1434 (2010).PubMed 

Google Scholar 
L. Kelley, A., J. Lunden, J., 1 Ocean Acidification Research Center, College of Fisheries and Ocean Sciences, University of Alaska, Fairbanks, Fairbanks, AK, 99775, USA, & 2 Haverford College, Haverford, PA, 19041, USA. Meta-analysis identifies metabolic sensitivities to ocean acidification. AIMS Environ. Sci. 4, 709–729 (2017).Stumpp, M. et al. Acidified seawater impacts sea urchin larvae pH regulatory systems relevant for calcification. Proc. Natl. Acad. Sci. U. S. A. 109, 18192–18197 (2012).ADS 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Stumpp, M. et al. Digestion in sea urchin larvae impaired under ocean acidification. Nat. Clim. Change 3, 1044–1049 (2013).ADS 
CAS 

Google Scholar 
Runcie, D. E. et al. Genomic characterization of the evolutionary potential of the sea urchin Strongylocentrotus droebachiensis facing ocean acidification. Genome Biol. Evol. 8, 272 (2017).
Google Scholar 
Sewell, M. Utilization of lipids during early development of the sea urchin Evechinus chloroticus. Mar. Ecol. Prog. Ser. 304, 133–142 (2005).ADS 
CAS 

Google Scholar 
Lucas, M. I., Walker, G., Holland, D. L. & Crisp, D. J. An energy budget for the free-swimming and metamorphosing larvae of Balanus balanoides (Crustacea: Cirripedia). Mar. Biol. 55, 221–229 (1979).
Google Scholar 
Shilling, F. M., Hoegh-Guldberg, O. & Manahan, D. T. Sources of energy for increased metabolic demand during metamorphosis of the abalone Haliotis rufescens (Mollusca). Biol. Bull. 191, 402–412 (1996).CAS 
PubMed 

Google Scholar 
Meidel, S. K. & Scheibling, R. E. Effects of food type and ration on reproductive maturation and growth of the sea urchin Strongylocentrotus droebachiensis. Mar. Biol. 134, 155–166 (1999).
Google Scholar 
Pearce, C. M. & Scheibling, R. E. Induction of metamorphosis of larvae of the green sea urchin, Strongylocentrotus droebachiensis by coralline red algae. Biol. Bull. 179, 304–311 (1990).CAS 
PubMed 

Google Scholar 
Gosselin, P. & Jangoux, M. From competent larva to exotrophic juvenile: a morphofunctional study of the perimetamorphic period of Paracentrotus lividus (Echinodermata, Echinoida). Zoomorphology 118, 31–43 (1998).
Google Scholar 
Hinegardner, R. T. Growth and development of the laboratory cultured sea urchin. Biol. Bull. 137, 465–475 (1969).CAS 
PubMed 

Google Scholar 
Strathmann, R. R. Length of pelagic period in echinoderms with feeding larvae from the Northeast Pacific. J. Exp. Biol. Ecol. 34, 23–27 (1978).
Google Scholar 
Byrne, M. et al. Unshelled abalone and corrupted urchins: Development of marine calcifiers in a changing ocean. Proc. Biol. Sci. 278, 2376–2383 (2011).PubMed 

Google Scholar 
Dupont, S., Dorey, N., Stumpp, M., Melzner, F. & Thorndyke, M. Long-term and trans-life-cycle effects of exposure to ocean acidification in the green sea urchin Strongylocentrotus droebachiensis. Mar. Biol. 160, 1835–1843 (2013).CAS 

Google Scholar 
Uthicke, S. et al. Impacts of ocean acidification on early life-history stages and settlement of the coral-eating sea star Acanthaster planci. PLoS ONE 8, e82938 (2013).ADS 
PubMed 
PubMed Central 

Google Scholar 
Dupont, S., Lundve, B. & Thorndyke, M. Near future ocean acidification increases growth rate of the lecithotrophic larvae and juveniles of the sea star Crossaster papposus. J. Exp. Zool. Mol. Dev. Evol. 314, 382–389 (2010).
Google Scholar 
Lim, Y.-K., Dang, X. & Thiyagarajan, V. Transgenerational responses to seawater pH in the edible oyster, with implications for the mariculture of the species under future ocean acidification. Sci. Total Environ. 782, 146704 (2021).ADS 
CAS 
PubMed 

Google Scholar 
Hettinger, A. et al. Persistent carry-over effects of planktonic exposure to ocean acidification in the Olympia oyster. Ecology 93, 2758–2768 (2012).PubMed 

Google Scholar 
Hettinger, A. et al. Larval carry-over effects from ocean acidification persist in the natural environment. Glob. Change Biol. https://doi.org/10.1111/gcb.12307 (2013).Article 

Google Scholar 
Albright, R. & Langdon, C. Ocean acidification impacts multiple early life history processes of the Caribbean coral Porites astreoides. Glob. Change Biol. 17, 2478–2487 (2011).ADS 

Google Scholar 
Yuan, X. et al. Elevated CO2 delays the early development of scleractinian coral Acropora gemmifera. Sci. Rep. 8, 2787 (2018).ADS 
PubMed 
PubMed Central 

Google Scholar 
Maboloc, E. A. & Chan, K. Y. K. Parental whole life cycle exposure modulates progeny responses to ocean acidification in slipper limpets. Glob. Change Biol. 2, 15647. https://doi.org/10.1111/gcb.15647 (2021).Article 

Google Scholar 
Mos, B., Byrne, M. & Dworjanyn, S. A. Effects of low and high pH on sea urchin settlement, implications for the use of alkali to counter the impacts of acidification. Aquaculture 528, 735618 (2020).CAS 

Google Scholar 
Harianto, J., Aldridge, J., Torres Gabarda, S. A., Grainger, R. J. & Byrne, M. Impacts of acclimation in warm-low pH conditions on the physiology of the sea urchin Heliocidaris erythrogramma and carryover effects for juvenile offspring. Front. Mar. Sci. 7, 588938 (2021).
Google Scholar 
Houlihan, E. P., Espinel-Velasco, N., Cornwall, C. E., Pilditch, C. A. & Lamare, M. D. Diffusive boundary layers and ocean acidification: Implications for sea urchin settlement and growth. Front. Mar. Sci. 7, 577562 (2020).
Google Scholar 
Norderhaug, K. M. & Christie, H. C. Sea urchin grazing and kelp re-vegetation in the NE Atlantic. Mar. Biol. Res. 5, 515–528 (2009).
Google Scholar 
Dickson, A., Sabine, C. L. & Christian, J. R. Guide to best practices for ocean CO2 measurements. (PICES Special Publication 3;191 pp, 2007).Lavigne, H. & Gattuso, J.-P. seacarb: seawater carbonate chemistry with R. R package version 2.4. http://CRAN.R-project.org/package=seacarb. (2011).R Core Team. R: A language and environment for statistical computing. R: A language and environment for statistical computing (2017).Guillard, R. R. L. & Ryther, J. H. Studies of marine planktonic diatoms. I. Cyclotella nana Hustedt and Detonula confervacea (Cleve) Gran. Can. J. Microbiol. 8, 229–239 (1962).CAS 
PubMed 

Google Scholar 
Stumpp, M., Wren, J., Melzner, F., Thorndyke, M. & Dupont, S. CO2 induced seawater acidification impacts sea urchin larval development I: Elevated metabolic rates decrease scope for growth and induce developmental delay. Comp. Biochem. Physiol. Mol. Integr. Physiol. 160, 331–340 (2011).CAS 

Google Scholar 
His, E., Heyvang, I., Geffard, O. & De Montaudouin, X. A comparison between oyster (Crassostrea gigas) and sea urchin (Paracentrotus lividus) larval bioassays for toxicological studies. Water Res. 33, 1706–1718 (1999).CAS 

Google Scholar 
U. S. National Institutes of Health, Bethesda, Maryland, U. ImageJ, Rasband, W.S., http://imagej.nih.gov/ij/.Smith, M. M., Cruz Smith, L., Cameron, R. A. & Urry, L. The larval stages of the sea urchin, Strongylocentrotus purpuratus. J. Morphol. 269, 713–733 (2008).PubMed 

Google Scholar 
Kahm, M., Hasenbrink, G., Lichtenberg-Frate, H., Ludwig, J. & Kschischo, M. grofit: Fitting Biological Growth Curves with R. J. Stat. Softw., 33(7), 1–21. URL http://www.jstatsoft.org/v33/i07/. (2010).Pinheiro, J., Bates, D., & R-core. Package ‘nlme’: Linear and Nonlinear Mixed Effects Models. Cran-R (2018).Pan, T.-C.F., Applebaum, S. L. & Manahan, D. T. Experimental ocean acidification alters the allocation of metabolic energy. Proc. Natl. Acad. Sci. U. S. A. 112, 4696–4701 (2015).ADS 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Jager, T., Ravagnan, E. & Dupont, S. Near-future ocean acidification impacts maintenance costs in sea-urchin larvae: Identification of stress factors and tipping points using a DEB modelling approach. J. Exp. Mar. Biol. Ecol. 474, 11–17 (2016).
Google Scholar 
Hoegh-Guldberg, O. & Emlet, R. B. Energy use during the development of a lecithotrophic and a planktotrophic echinoid. Biol. Bull. 192, 27–40 (1997).CAS 
PubMed 

Google Scholar 
Vaïtilingon, D. et al. Effects of delayed metamorphosis and food rations on the perimetamorphic events in the echinoid Paracentrotus lividus (Lamarck, 1816) (Echinodermata). J. Exp. Mar. Biol. Ecol. 262, 41–60 (2001).
Google Scholar 
García, E., Clemente, S. & Hernández, J. C. Ocean warming ameliorates the negative effects of ocean acidification on Paracentrotus lividus larval development and settlement. Mar. Environ. Res. 110, 61–68 (2015).PubMed 

Google Scholar 
Wangensteen, O. S., Dupont, S., Casties, I., Turon, X. & Palacín, C. Some like it hot: Temperature and pH modulate larval development and settlement of the sea urchin Arbacia lixula. J. Exp. Mar. Biol. Ecol. 449, 304–311 (2013).
Google Scholar 
García, E., Clemente, S. & Hernández, J. C. Effects of natural current pH variability on the sea urchin Paracentrotus lividus larvae development and settlement. Mar. Environ. Res. 139, 11–18 (2018).PubMed 

Google Scholar 
Marshall, D. J. & Keough, M. J. Variation in the dispersal potential of non-feeding invertebrate larvae: The desperate larva hypothesis and larval size. Mar. Ecol. Prog. Ser. 255, 145–153 (2003).ADS 

Google Scholar 
Huggett, M. J., Williamson, J. E., de Nys, R., Kjelleberg, S. & Steinberg, P. D. Larval settlement of the common Australian sea urchin Heliocidaris erythrogramma in response to bacteria from the surface of coralline algae. Oecologia 149, 604–619 (2006).ADS 
PubMed 

Google Scholar 
Espinel-Velasco, N., Agüera, A. & Lamare, M. Sea urchin larvae show resilience to ocean acidification at the time of settlement and metamorphosis. Mar. Environ. Res. 159, 104977 (2020).CAS 
PubMed 

Google Scholar 
Lamare, M. & Barker, M. Settlement and recruitment of the New Zealand sea urchin Evechinus chloroticus. Mar. Ecol. Prog. Ser. 218, 153–166 (2001).ADS 

Google Scholar 
Martin, S. et al. Early development and molecular plasticity in the Mediterranean sea urchin Paracentrotus lividus exposed to CO2-driven acidification. J. Exp. Biol. 214, 1357–1368 (2011).CAS 
PubMed 

Google Scholar 
Vargas, C. A. et al. Species-specific responses to ocean acidification should account for local adaptation and adaptive plasticity. Nat. Ecol. Evol. 1, 0084 (2017).
Google Scholar 
Espinel-Velasco, N. et al. Effects of ocean acidification on the settlement and metamorphosis of marine invertebrate and fish larvae: a review. Mar. Ecol. Prog. Ser. 606, 237–257 (2018).ADS 

Google Scholar 
Briffa, M., de la Haye, K. & Munday, P. L. High CO2 and marine animal behaviour: potential mechanisms and ecological consequences. Mar. Pollut. Bull. 64, 1519–1528 (2012).CAS 
PubMed 

Google Scholar 
Gaylord, B. et al. Ocean acidification through the lens of ecological theory. Ecology 96, 3–15 (2015).PubMed 

Google Scholar  More

Ebola datasetThe 2018–2020 DRC EVD outbreak lasted over 24 months and spread over 3 distinct spatial and temporal waves. Between the emergency declaration of the EVD outbreak in northern DRC on August 1, 2018 and the outbreak’s official end on June 25, 2020, the DRC Ministry of Health has reported a total of 3481 cases (including confirmed and probable), 1162 recoveries, and 2299 deaths16 in the provinces of Northern Kivu, Southern Kivu, and Ituri. The dataset considered here is a large subset of the entire EVD database compiled by the University of Kinshasa School of Public Health, which comprises 3117 total case records (confirmed and probable) recorded between May 3, 2018, and September 12, 2019. The data included partially de-identified but still detailed patient information, such as each person’s location, date of symptom onset and hospitalization, as well as discharge due to recovery or death. These individual records came from the Ebola treatment centers in 24 different health zones, spread out among the three DRC provinces of Northern Kivu, Southern Kivu, and Ituri.Of the 24 health zones, 77.1% of all cases were from only 6: Beni, Butembo, Katwa, Kalunguta, Mabalako, and Mandima. Only 9.7% of cases were under the age of 18. There is also a slightly larger proportion of females contracting the disease, comprising 57.0% of the cases. Approximately 5% of the cases were health care workers. About one-third of the EVD fatalities were not identified until patient’s death and thus not effectively isolated from the time of infection. Although over 170,000 contacts of confirmed and probable Ebola cases had been monitored across all affected health zones for 21 days after their last known exposure by the end of the epidemic, some of the contact tracing was incomplete due to insecurity that prevented public health response teams from entering some communities. The overall case density map is presented in panel (A) of Fig. 1 with the animated version of the map presented in the online appendix in Fig. A.1. Notice that the high-density areas, particularly Butembo, Katwa, and Beni, are all spatially small health zones corresponding to cities or towns with larger populations.Figure 1DRC Ebola dataset. (A) The spatial distribution of 3481 EVD cases across the northern DRC health zones during Ebola 2018–2020 outbreak. (B) The flowchart of personal records available up to September 12, 2019 available for the current analysis. The total number of available individual disease records was 3080. Map created using open software R17 with geospatial data obtained from18.Full size imageFigure 2Daily incidence and removal rates. Daily incidence (grey bars) and removal counts (red dots) during DRC Ebola 2018–2020 outbreak between August 15, 2018 and September 12, 2020 along with their respective trendlines (loess smoothers). The blue trendline above the plot represents daily effective reproduction number (mathcal{R}_t) defined as the ratio of daily number of new infections to new removals. The vertical lines indicate cut-off dates for data collection in each wave as listed in Table 1.Full size imageTable 1 Observed cases by EVD wave.Full size tableCase alerts and definitionsSince early August, 2018, the DRC Ministry of Health has been collaborating with several international partners to support and enhance EVD response activities through its emergency operations center in Goma. To the extent possible given regional security considerations19, the response teams were deployed to interview patients and their suspected contacts using a standardized case investigation form classifying cases as suspected, probable, or confirmed. A suspected case (whether surviving or not) was defined as one with the acute onset of fever (over 100(^{circ })F) and at least three Ebola-compatible clinical signs or symptoms (headache, vomiting, anorexia, diarrhea, lethargy, stomach pain, muscle or joint aches, difficulty swallowing or breathing, hiccups, unexplained bleeding, or any sudden, unexplained death) in a North Kivu, South Kivu, or Ituri resident or any person who had traveled to these provinces during this period and reported the signs or symptoms defined above. A patient who met the suspected case definition and died but from whom no specimens were available was considered a probable case. A confirmed Ebola case was defined as a suspected case with at least one positive test for Ebola virus using reverse transcription polymerase chain reaction (RT-PCR)20 testing. Patients with suspected Ebola were isolated and transported to an Ebola treatment center for confirmatory testing and treatment2.Onset and removalIn our analysis of the DRC dataset, we focused on dates of symptom onset and removal, with removal defined as either a death/recovery at home or transfer to an Ebola treatment center (ETC). It was assumed that, once in the treatment center, the probability of further infection spread by an isolated individual was very small due to the strict safety protocols—and later due also to vaccination of healthcare personnel and family members who were in contact with the suspected Ebola case. As summarized in panel (B) of Fig. 1, we were able to access 3117 out of 3481 individual records of confirmed and probable Ebola cases. Of these 3117 records, 37 were missing both the onset and recovery dates and were removed from further analysis. In about 30% of the remaining records, either their dates of onset or removal were missing. A detailed flow diagram summarizing the amount of missing data and data processing leading to the final dataset is presented in panel (B) of Fig. 1. The distribution of the original and the partially imputed records across the three waves of infection is provided for further reference in Table 1.Spatial and temporal patternsThroughout the pandemic, the incidence rates exhibited strong spatial and temporal patterns that can be summarized as three distinct waves of infections with approximate boundaries marked by vertical lines in Fig. 1. The distribution of weekly reported cases across the most affected health zones listed in Table 1 is provided in the bar plot and in the corresponding animation in the appendix (see Figure A.1). As seen from the bar chart and the animated plot, the epidemic was initially driven largely by infections in the health zones of Beni, Mandima and Mabalako. After several months, the incidence of new cases in these zones subsided, but the epidemic moved south to the health zones of Katwa and Butembo, where the majority of new infections was registered between weeks 22 to 45 of the epidemic (see Panel (A) in Figure A.1 in the online Appendix). In the final spatial shift, around week 49, the epidemic returned to the health zones of Beni, Mandima, and Mabalako, where it was mostly extinguished around week 60 (September 2019). Isolated Ebola incidences occurred sporadically across northern DRC until end of the outbreak was officially declared in June 2020.The empirical patterns of incidence and removal for EVD cases are summarized in Fig. 2 with the bar and the dot plots representing the daily numbers of new infections and removals, respectively. As seen from the plot, these daily counts closely follow a three-wave temporal pattern in Table 1. This is further evident from the black and red trendlines representing the loess smoothers (see21). The daily ratio of new cases and removals may be interpreted as a crude estimate of the effective reproduction number (mathcal{R}_t) defined more formally in (2) in Model for Data Analysis below. In particular, the blue trendline for (mathcal{R}_t) indicates that towards the end of the observed time period, the number of removals outpaced the number of new infections ((mathcal{R}_t 0) and (r_t = 0) where (beta > 0) is the rate of infection, (gamma > 0) is the rate of recovery and (rho > 0) is the initial amount of infection. In particular, the model implies the existence of the basic reproduction number (mathcal{R}_0) (R-naught), which determines the average speed of disease spread11 and is given by the formula$$mathcal{R}_0=beta /gamma .$$If (mathcal{R}_0 > 1), the proportion of infected initially rises and then subsides, with the final proposition of surviving susceptibles given by (s_infty = 1 – tau > 0) where (tau) is know as the epidemic’s final size. In typical statistical analysis, an estimate of (mathcal{R}_0) is obtained by separately estimating the parameters (beta) and (gamma). Another important quantity related to (1) is the effective reproduction number, which is typically defined as$$begin{aligned} mathcal{R}_t= mathcal{R}_0 s_t. end{aligned}$$
(2)
Although equation (1) is typically considered in the context of an average behavior of a large population, for our purposes we interpret it as defining the individual histories of infection and recovery, according to the idea of the dynamic survival analysis (DSA) discussed recently in10 and24 and also briefly summarized in the Appendix. With the DSA approach, we interpret equation (1) as the so-called stochastic master equation25 describing the change in probability of a randomly selected individual being at time t either susceptible, infected, or removed. These respective probabilities are represented by the scaled proportions (s_t/(1+rho )), (iota _t/(1+rho )), and (r_t/(1+rho )) and evolve according to (1). As outlined in10, the DSA-based interpretation of the classical SIR equations has a number of advantages that make it particularly convenient for analyzing epidemic data consisting of individual histories of infection onsets and removals, which is exactly the type of data available in the DRC Ebola dataset. The fact that the model is individual-based implies also that we can vary the parameters (theta =(beta ,gamma ,rho )) to account for individual covariates and changes in the parameter values over time, as different waves of infection sweep through the population. Finally, for the purpose of our analysis, it is also important to note that the DSA model does not require any knowledge of the size of the susceptible population subjected to the epidemic pressure. For the DRC dataset, that assumption would be difficult to justify due to spatial and temporal heterogeneity of the epidemic and the frequent movements of local populations driven by political conflicts and insecurity. Another element complicating the determination of the size of susceptible population was the ring vaccination campaign that has been conducted since 2019 wherever possible in the northern DRC during periods of relative stability, despite local mistrust and supply issues. This campaign ultimately resulted in over 250,000 vaccinations.Note that, because (s_0 = 1), the values of (mathcal{R}_0) and (mathcal{R}_t) coincide for (t = 0). Moreover, (s_t = exp left( -mathcal{R}_0 int _0^t r_u mathrm {d}u right)) is a decreasing function of time and therefore, so is (mathcal{R}_t). However, in practice, this implication is problematic. Rewriting (mathcal{R}_t = – {dot{s}}_t/ {dot{r}}_t) suggests that a crude but sensible way to estimate (mathcal{R}_t) empirically is to take the ratio of daily number of new infections to new removals. The empirical (mathcal{R}_t) thus estimated will not be necessarily monotonically decreasing. In the light of possibly changing parameters and the effective population size, we have adopted this approach to estimating the daily effective reproduction number (mathcal{R}_t) in Fig. 2.Parameter estimationWe assume that, for each of the three waves of the epidemic, we have a separate and independent set of parameters (theta) and that, in each wave, we observe (n_T) histories (records) of infection. The i-th individual history may be represented either by the times of disease onset and removal ((t_i,T_i)) or by (t_i) or (T_i) times alone ((t_i,circ )) or ((circ ,T_i)) ((circ) denoting missing value). We assume that among the available (n_T) histories we have n complete records ((t_i,T_i)), (n_1) incomplete ones ((t_i,circ )) and (n_2) incomplete ones ((circ ,T_i )). The wave-specific DSA likelihood function for n complete data records is (see Appendix)$$begin{aligned} begin{aligned} {mathcal {L}}_C(theta vert t_1ldots ,t_n,T_1,ldots ,T_n,T)=(s_T-1)^{-n}prod _{i=1}^n {dot{s}}_{t_i}gamma ^{w_i}e^{-gamma (T_i wedge T -t_i)} end{aligned} end{aligned}$$
(3)
where T is the available time horizon and (w_i) is the binary variable indicating whether (T_i) is right-censored (that is, (T_iwedge T =T)) in which case (w_i = 0) and otherwise (w_i = 1). For the remaining (n_1+n_2) records that are partially incomplete, the wave-specific DSA likelihood function is$$begin{aligned} begin{aligned} {mathcal {L}}_I(theta vert t_1ldots ,t_{n_1},T_1,ldots ,T_{n_2},T)= (s_T-1)^{-(n_1+n_2)} gamma ^{n_2}prod _{i=1}^{n_1} {dot{s}}_{t_i} prod _{i=1}^{n_2} (rho e^{-gamma T_i }-iota _{T_i}) end{aligned} end{aligned}$$
(4)
where we assume that (T_i1). Given the wave-specific time horizons (T’s), the set of parameters for each epidemic wave was estimated independently using 2 independent chains of 3000 iterations, with a burn-in period of 1000 iterations. The chains’ convergence assessed using Rubin’s R statistic28. The analysis resulted in approximate samples from the posterior distribution of (theta) for each of the three waves of the epidemic (see e.g., Fig. 4).Ethics statement on human subjects and methodsThe research was conducted in accordance with the relevant guidelines and regulations of the US law and OSU Institutional Review Board. The research activities involving human subjects discussed in the paper meet the US federal exemption criteria under 45 CFR 46 and 21 CFR 56. More

To explore how environmental gradients shape the distribution of cyanophages and picocyanobacteria, we conducted high-resolution surveys in surface waters along five oceanic transects on three cruises covering thousands of kilometres in the North Pacific Ocean in the spring or early summer of 2015, 2016 and 2017 (Fig. 1a–c). These cruises, two of which were out-and-back, passed through distinct regimes from warm, saline and nutrient-poor waters of the North Pacific Subtropical Gyre to cooler, less saline and nutrient-rich waters of higher latitudes influenced by the subpolar gyre (Fig. 1d–i)27. The shift between the two gyres was marked by abrupt changes in trophic indicators such as particulate carbon concentrations (Fig. 1g) and a chlorophyll front (defined as the 0.2 mg m−3 chlorophyll contour28; Fig. 1a–c). As such, the inter-gyre transition zone, defined by salinity and temperature thresholds29 (Fig. 1d), was distinct from both the subtropical and subpolar gyre ecosystems28.Fig. 1: Gradients in environmental conditions across the North Pacific gyres.a–c, Transects of three cruises overlaid on monthly averaged satellite-derived sea-surface chlorophyll in March 2015 (a), April 2016 (b) and June 2017 (c). d, Temperature–salinity diagram showing the boundaries of the subtropical and subpolar gyres (black dashed lines) based on the salinity thresholds reported by Roden29. e–i, Temperature (e), salinity (f) as well as the levels of particulate carbon (g), phosphate (h) and nitrate + nitrite (i) as a function of latitude. The coloured dashed lines show the position of the 0.2 mg m−3 chlorophyll contour. For environmental variables plotted against temperature, see Supplementary Fig. 3.Full size imageUnexpected Prochlorococcus declineProchlorococcus concentrations in the oligotrophic waters of the subtropical gyre were 1.5–3.0 × 105 cells ml−1, comprising an average of approximately 29% of the total bacteria (Extended Data Fig. 1) and numerically dominating the phytoplankton community in all three cruises (Extended Data Fig. 2). Prochlorococcus abundance remained high in the southern region of the transition zone in 2015 and 2016, decreasing precipitously to less than 2,000 cells ml−1 north of the chlorophyll front, generally constituting 80% of cyanophages measured, with the remainder consisting of T7-like clade A and TIM5-like cyanophages (Fig. 3 and Extended Data Fig. 4). Cyanophage abundances correlated positively with total picocyanobacteria in the subtropical gyre (Pearson’s coefficient of multiple correlation (r) = 0.54, P = 0.02, n = 26; Fig. 2d), suggesting that cyanophages were limited by the availability of susceptible hosts in this region and were not regulating picocyanobacterial populations. On average, less than 1% of the cyanobacterial populations were infected (Fig. 4), with higher infection rates by T4-like cyanophages than T7-like cyanophages (Extended Data Figs. 5 and 6). These instantaneous measurements of infection were used to estimate the daily rates of mortality39 (Methods and Supplementary Discussion), which suggests that 0.5–6% of picocyanobacterial populations were lysed by viruses each day (Extended Data Fig. 7). This implicates other factors, such as grazing45, as the major causes of cyanobacterial mortality in the North Pacific Subtropical Gyre.Fig. 3: Cyanophage community composition across the North Pacific gyres.a–c, Cyanophage abundance for the March 2015 (a), April 2016 (b) and June 2017 (c) transects. Insets: T7-like clade A and TIM5-like cyanophage abundances on an expanded scale (similar to the main images, the units for the vertical axes are ×105 viruses ml−1). The grey shaded regions show the position of the virus hotspot. See Extended Data Fig. 4 for the confidence intervals and out-and-back reproducibility and Supplementary Fig. 4 for cyanophage lineages plotted against latitude.Full size imageFig. 4: Viral infection patterns of picocyanobacteria in the North Pacific Ocean.a–f, Viral infection levels (black) of Prochlorococcus (a,c,e) and Synechococcus (b,d,f) plotted against temperature for the March 2015 (a,b), April 2016 (c,d) and June 2017 (e,f) transects. Insets: infection levels on an expanded scale. The solid lines show infection (red), Prochlorococcus (green) and Synechococcus (pink) averaged and plotted for every 0.5 °C. The dashed lines and shaded regions show the position of the chlorophyll front and the virus hotspot, respectively. For plots by latitude and the upper and lower bounds of infection, see Extended Data Figs. 5 and 6.Full size imageWithin the transition zone we observed a steep latitudinal increase in the abundance of cyanophages for every transect, which we define as a cyanophage hotspot (Fig. 2c and Extended Data Figs. 2 and 4). The cyanophage abundances in this hotspot were between three- and tenfold greater than in the subtropical gyre (Fig. 2c). Notably, cyanophages were approximately 25% more abundant (an increase of approximately 5 × 105 viruses ml−1) in the hotspot on the 2017 cruise relative to the other two cruises, reaching a maximum of 2 × 106 viruses ml−1. The hotspot peaked at temperatures of 15–16 °C on all transects, regardless of the geographical location, season or the exact pattern of the Prochlorococcus and Synechococcus distributions (Fig. 2c). Notably, the numbers of T7-like clade B cyanophages increased sharply in the transition zone to become the most abundant lineage, whereas T4-like cyanophages increased more modestly (Fig. 3 and Extended Data Fig. 4). The change in the cyanophage community structure was particularly pronounced in June 2017, when T7-like cyanophages were up to 2.3-fold more abundant than T4-like cyanophages (Fig. 3c). The switch in the relative abundance of T4-like and T7-like clade B cyanophages was diagnostic of the cyanophage hotspot compared with patterns in the subtropical and subpolar gyres.To begin assessing whether cyanophages negatively affected cyanobacterial populations in the hotspot, we tested the relationship between the abundance of cyanophages and total cyanobacteria. This showed a significant negative correlation between cyanophage and cyanobacterial abundances across all three cruises (Pearson’s r = −0.56, two-sided P = 0.0005, n = 34). This relationship was particularly distinct in 2017, when cyanobacteria were at their overall lowest abundances and cyanophages at their highest (Pearson’s r = −0.65, two-sided P = 0.004, n = 18). This suggests that viruses are one of the key regulators of picocyanobacteria in the region of the hotspot. However, no significant correlation was found across all regimes and all years (Pearson’s r = −0.008, two-sided P = 0.9, n = 87; Fig. 2d), indicating that factors other than viruses are likely to be more important in regulating the abundances of cyanobacteria in other regimes.Our single-cell infection measurements allowed us to directly evaluate active viral infection and its impact on picocyanobacteria in the transition zone. Viral infection spiked in this region each year with infection levels that were an average of two- to ninefold higher than those in the subtropical gyre (Fig. 4 and Extended Data Figs. 5,6 and 8). Infection peaked within the temperature range of 12–18 °C and was associated with a concomitant dip in Prochlorococcus abundances in all three cruises (Fig. 4 and Extended Data Fig. 5). These findings provide independent support for the strong negative correlation between cell and virus abundances (Fig. 2d) being the result of virus-induced mortality.Lineage-specific infection was also distinct in the transition zone relative to the subtropical gyre. Infection by T7-like clade B cyanophages generally increased to reach (2015 and 2016) or exceed (2017) those of T4-like cyanophages (Extended Data Figs. 5 and 6). In addition, the ratio of the abundances of T7-like clade B cyanophages to the number of cells they infected was 2.6-fold greater in the hotspot than the subtropics, whereas this ratio was similar in both regions for T4-like cyanophages. Together, these results indicate that, within the hotspot, the T4-like cyanophages displayed increased levels of infection, whereas the T7-like cyanophages displayed both increased levels of infection and produced more viruses per infection, suggesting that T7-like clade B cyanophages are better adapted to conditions in the transition zone (see below).Of the three cruises, the highest levels of viral infection were observed in June 2017, with up to 9.5% and 8.9% of Prochlorococcus and Synechococcus infected, respectively (Fig. 4e,f). This dramatic increase in infection mirrored the massive decline in Prochlorococcus abundances (Fig. 4e and Extended Data Fig. 5i). We estimate that viruses killed 10–30% of Prochlorococcus and Synechococcus cells daily at these high instantaneous levels of infection (Extended Data Fig. 6) based on the expected number of infection cycles cyanophages were able to complete at the light and temperature conditions in the transition zone (Methods and Supplementary Discussion). Given that Prochlorococcus is estimated to double every 2.8 ± 0.8 d at the low temperatures in this region12, we estimate that 21–51% of the population was infected and killed in the interval before cell division. Synechococcus is expected to have faster growth rates at these temperatures, doubling every 1.1 ± 0.2 d (refs. 12,46). Thus, we estimate that less of the Synechococcus population (9–31%) was killed before division.Under quasi-steady state conditions, abiotic controls on the growth rate of Prochlorococcus are balanced by mortality due to viral lysis, grazing and other mortality agents39,45,47. Based on the high levels of virus-mediated mortality, the parallel pattern between Prochlorococcus’ death and viral infection, and the negative correlation between cyanophage and picocyanobacterial abundances in the transition zone, we propose that enhanced viral infection in 2017 disrupted this balance, leading to the unexpected decline in Prochlorococcus populations. Grazing and other mortality agents not investigated here could also have contributed to additional mortality beyond the steady state, resulting in further losses of Prochlorococcus. In contrast to Prochlorococcus, Synechococcus maintained large populations despite high levels of infection (Fig. 4f), presumably due to their faster growth rates enabling them to maintain a positive net growth despite enhanced mortality. These findings suggest that virus-mediated mortality in 2017 was an important factor in limiting the geographic range of Prochlorococcus that resulted in a massive loss of habitat of approximately 550 km.Cyanophage abundances and infection levels dropped sharply in the higher-latitude waters north of the hotspot (Figs. 2c, 4 and Extended Data Figs. 1d,h and 2). The abundances of both T7-like clade B and T4-like cyanophages declined precipitously, yet T4-like cyanophages were the dominant cyanophage lineage (Fig. 3). T7-like clade A cyanophages generally increased locally at the northern border of the hotspot and became the dominant T7-like lineage in two samples between 38 and 39.2° N in 2017 (Fig. 3c and Extended Data Fig. 4). In contrast to all other cyanophages, the abundances of TIM5-like cyanophages increased in waters north of the hotspot (Fig. 3 and Extended Data Fig. 4d,i,m) but remained a minor component of the cyanophage community. No relationship was found between cyanophage and cyanobacterial abundances (Fig. 2d), and less than 1.5% of picocyanobacteria were infected by all cyanophage lineages in these waters (Fig. 4).The cyanophage hotspot in the transition zone is a ridge of high virus activity that separates the subtropical and subpolar gyres. The reproducibility of our observations, which were separated by days to weeks within each cruise (2016 and 2017) and by years among the three cruises (Extended Data Fig. 4), indicates that this virus hotspot is a recurrent feature at the boundary of these two major gyres in the North Pacific Ocean. This suggests that the hotspot forms due to the distinctive environment of the inter-gyre transition zone creating conditions that enhance infection of picocyanobacteria and proliferation of cyanophages. Prochlorococcus in the transition zone may be prone to stress due to being close to the limits of their temperature growth range5,6, which has the potential to increase susceptibility to viral infection. Alternatively, there may be temperature-dependent trade-offs between virus decay and production that lead to replication optima within a narrow temperature range48. Cyanophage infectivity has been observed to decay more slowly at colder temperatures49, which may allow for the accumulation of infective viruses, leading to increased infection. In addition, cyanophage infections may be more productive due to enhanced nutrient supply in the transition zone27 (Fig. 1h,i) relative to the subtropics, given that the cyanophages replicate in hosts with presumably greater intracellular nutrient quota and obtain more extracellular nutrients, both of which may increase progeny production9,10. The environmental factors influencing the production and removal of viruses probably vary in intensity at different times, leading to variability in cyanophage abundance and infection levels. Thus, the putative cyanophage replication optimum in the hotspot may reflect the combined effects of temperature and nutrient conditions that are intrinsically linked to the oceanographic forces that shape the transition zone itself.Changes in the cyanophage community structure over environmental gradients are likely to reflect differences in host range, infection properties and genomic potential to remodel host metabolism9. Our data, together with previous measurements in the North Pacific Subtropical Gyre38,39, indicate that the T4-like cyanophages are the lineage best adapted to the low-nutrient waters of the subtropics (Fig. 2d–f). As these waters are inhabited by hundreds of genomically diverse subpopulations of Prochlorococcus50, the broad host range of many T4-like cyanophages18,19,22,51 may be advantageous for finding a suitable host. T4-like cyanophages also have a large and diverse repertoire of host-derived genes21,51—such as nutrient acquisition, photosynthesis and carbon-metabolism genes—that augment host metabolism52 and may increase fitness in nutrient-poor conditions in the subtropics51. In contrast, T7-like clade A and B cyanophages seem to be better adapted to conditions in the transition zone (Fig. 3). T7-like cyanophages have narrow host ranges19,22,40, with smaller genomes and fewer genes to manipulate the host metabolism23, which may allow them to replicate and produce more progeny in regions with elevated nutrient concentrations relative to subtropical conditions. The maximal abundances of TIM5-like cyanophages were found in the most productive waters at the northern end of the transects where the cyanobacterial abundances were lowest and Synechococcus was the dominant picocyanobacterium. This may be partially due to the narrow host range of TIM5-like cyanophages and their specificity for Synechococcus40,44. Our findings of reproducible lineage-specific responses to changing ocean regimes indicate that cyanophage lineages occupy distinct ecological niches.Temperature and nutrient changes occurring in the transition zone are expected to result in shifts in picocyanobacterial diversity at the sub-genus level (Supplementary Discussion), which we speculate may affect community susceptibility to viral infection. One mechanism for this may be that the picocyanobacteria that thrive in the transition zone are intrinsically more susceptible to viral infection. Another scenario may be related to trade-offs associated with the evolution of resistance to viral infection. The horizontal advection of nutrient-rich waters to the transition zone28 may select for rapidly growing cells adapted for efficient resource utilization. Viral resistance in picocyanobacteria often incurs the cost of reduced growth rates53,54. Thus, competition for nutrients in this region may favour cells with faster growth rates but increased susceptibility to viral infection. Thus, it is probable that the cyanophage distributions do not always follow the cyanobacterial patterns (Extended Data Fig. 2) because of complex interactions between lineage-specific cyanophage traits, host community structure and environmental variables, which may vary seasonally or annually as a result of interannual variability in environmental conditions (see below).Despite consistent features in cyanophage distributions across the North Pacific Ocean, cyanophage infection was higher (Fig. 4 and Extended Data Fig. 7), whereas Prochlorococcus abundances were consistently lower (Fig. 2a), across the June 2017 transects relative to the March 2015 and April 2016 transects. Seasonality and/or climate variability could explain this interannual variability, although the data currently available to assess this are sparse. Viral infection of picocyanobacteria in the subtropical gyre increased from early spring to summer, suggesting a potential seasonal pattern that may extend across the transect (Extended Data Fig. 9a). In addition, the June 2017 transect occurred during a neutral-to-negative El Niño phase with lower sea-surface temperatures relative to the 2015 and 2016 transects, which were in years of a record marine heatwave, followed by a strong El Niño55 (Extended Data Fig. 9b). In 2015 and 2016, the Prochlorococcus abundances were found to be higher than usual in the North Pacific Ocean in this (Fig. 2a) and other studies56,57. Irrespective of the underlying drivers for the observed interannual variability, we speculate that an ecosystem tipping point was reached in the hotspot under the prevailing conditions in June 2017, aided by the higher cyanophage abundances yet smaller Prochlorococcus population sizes. In this scenario, picocyanobacterial populations were subjected to high infection levels that resulted in an accumulation of cyanophages, initiating a stronger than usual positive-feedback loop between infection and virus production, and precipitating the unexpected Prochlorococcus decline. Continued observations in the North Pacific Ocean are needed to evaluate the potential link between seasonality and/or large-scale climate forcing as ultimate drivers affecting virus–host interactions.Predicting basin-scale virus dynamicsMeasurements of cyanobacterial and cyanophage abundances rely on discrete sample collection from shipboard oceanographic expeditions, which limits the geographical and seasonal extent of available data. Therefore, we developed a multiple regression model based on high-resolution satellite data of temperature and chlorophyll to predict cyanophage abundances, a key proxy of cyanobacterial infection (Pearson’s r = 0.61, two-sided P = 1.7 × 10−8, degrees of freedom = 68, n = 70). We used the model to estimate the geographical extent of the virus hotspot. The model accurately predicted the location of the hotspot and cyanophage abundances along a fourth transect in April 2019 (Supplementary Table 1), with the majority of observations falling within the 95% confidence intervals of the model predictions (Fig. 5a–c). Application of the model to the larger region predicted that the virus hotspot formed a boundary extending across the North Pacific Ocean, with lower cyanophage abundances on both sides (Fig. 5d,e and Supplementary Fig. 1). This boundary had the hallmarks of the hotspot with a core that was dominated by T7-like cyanophages and the flanking gyre regions dominated by T4-like cyanophages. Thus, this feature may be more appropriately termed a ‘hot-zone’ due to its substantial projected aerial extent. Assuming the infection levels observed in the hotspot in June 2017 were similar throughout the hot-zone, the potential habitat loss for Prochlorococcus would be about 3.2 × 106 km2, approximately half of the cumulative area loss of the Amazonian rainforest to date58.Fig. 5: Prediction of cyanophage abundances.a–c, Model-based predictions of cyanophage abundances corresponding to the empirically measured total (a), T4-like (b) and T7-like clade B (c) cyanophage abundances along a transect in the North Pacific in April 2019. The shaded regions show the 95% confidence interval for the model predictions. d,e, Predicted total cyanophages (d) and the ratio of T4-like/T7-like clade B cyanophages (e) in June 2017 in the North Pacific Ocean. The black lines indicate the cruise track. The grey areas represent regions with no values due to cloud cover or that were beyond the limits of the predictive model. The hotspot peak corresponds to yellow regions in d and red regions in e.Full size imageVirus hotspot biogeochemistryWith the ability to predict biogeographic patterns of cyanophages, we evaluated the potential biogeochemical implications of virus-mediated picocyanobacterial lysis and release of organic material in sustaining the bacterial community6,7,8,9. The aerial extent of the hot-zone (approximately 4 × 106 km2) is only 14% of the size of the subtropical gyre (2.9 × 107 km2), and yet the total virus-mediated organic matter released from picocyanobacteria in the hot-zone in June 2017 was estimated to be on par with that for the entire North Pacific Subtropical Gyre (Methods and Supplementary Discussion). We estimate that viral lysate released from picocyanobacteria in the subtropical gyre could sustain 4.4 ± 0.8% of the calculated bacterial carbon demand there (Extended Data Fig. 10). In contrast, viral lysate released in the transition zone could sustain an average of 21 ± 12% of the bacterial carbon demand, reaching 33% in some regions (Extended Data Fig. 10), assuming that the bacterial assimilation and growth efficiencies were similar between the subtropical gyre and the hotspot. Thus, local generation of cyanobacterial viral lysate in the transition zone is likely to be an important source of carbon for the heterotrophic bacterial community that can rapidly utilize large molecular weight dissolved organic matter59 and may have contributed to the increase in their abundances south of the chlorophyll front in 2017 (Extended Data Fig. 1a,e). More