Ecology

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

La Sorte, F. A., Johnston, A. & Ault, T. R. Global trends in the frequency and duration of temperature extremes. Clim. Change 166, 1–2 (2021).Article 
ADS 

Google Scholar 
Hansen, J., Sato, M., Ruedy, R., Lo, K. & Medina-Elizade, M. Global temperature change. Proc. Natl. Acad. Sci. U.S.A. 103(39), 14288–14293 (2006).Article 
ADS 
CAS 

Google Scholar 
Borchert, R., Robertson, K., Schwartz, M. D. & Williams-Linera, G. Phenology of temperate trees in tropical climates. Int. J. Biometeorol. 50, 57–65 (2005).Article 
ADS 

Google Scholar 
Misra, G., Sarah, A. & Menzel, A. Ground and satellite phenology in alpine forests are becoming more heterogeneous across higher elevations with warming. Agric. For. Meteorol. 303, 108383 (2021).Article 
ADS 

Google Scholar 
Zuo, Z., Xiao, D. & Qiong, H. Role of the warming trend in global land surface air temperature variations. Sci. China Earth Sci. 6, 866–871 (2021).Article 
ADS 

Google Scholar 
Ling, Y. et al. Assessing the accuracy of forest phenological extraction from sentinel-1 C-band backscatter measurements in deciduous and coniferous forests. Remote Sens. 14(3), 674 (2022).Article 
ADS 

Google Scholar 
Zhang, H., Yuan, W., Liu, S., Dong, W. & Fu, Y. Sensitivity of flowering phenology to changing temperature in China. J. Geophys. Res. Biogeosci. 120(8), 1658–1665 (2015).Article 

Google Scholar 
Cho, J. G. et al. Apple phenology occurs earlier across South Korea with higher temperatures and increased precipitation. Int. J. Biometeorol. 65, 265–276 (2020).Article 

Google Scholar 
Li, C. et al. Response of vegetation phenology to the interaction of temperature and precipitation changes in Qilian mountains. Remote Sens. 14(5), 1248 (2022).Article 
ADS 

Google Scholar 
Berra, E. F. & Gaulton, R. Remote sensing of temperate and boreal forest phenology: A review of progress, challenges and opportunities in the intercomparison of in-situ and satellite phenological metrics. For. Ecol. Manage. 480, 118663 (2021).Article 

Google Scholar 
Zhang, Y. & Li, M. A new method for monitoring start of season (SOS) of forest based on multisource remote sensing. Int. J. Appl. Earth Obs. Geoinf. 104, 102556 (2021).
Google Scholar 
Zhang, X. et al. Monitoring vegetation phenology using MODIS. Remote Sens. Environ. 84(3), 471–475 (2003).Article 
ADS 

Google Scholar 
Thapa, S., Garcia Millan, V. E. & Eklundh, L. Assessing forest phenology: A multi-scale comparison of near-surface (UAV, spectral reflectance sensor, PhenoCam) and Satellite (MODIS, Sentinel-2) remote sensing. Remote Sens. 13, 1597 (2021).Article 
ADS 

Google Scholar 
Bórnez, K., Descals, A., Verger, A. & Peñuelas, J. Land surface phenology from VEGETATION and PROBA-V data: Assessment over deciduous forests. Int. J. Appl. Earth Observ. Geoinf. 84, 101974 (2020).
Google Scholar 
Yu, L., Yan, Z. & Zhang, S. Forest phenology shifts in response to climate change over China–Mongolia–Russia international economic corridor. Forests 11, 757 (2020).Article 

Google Scholar 
Lara, C. et al. Climatic regulation of vegetation phenology in protected areas along Western South America. Remote Sens. 13, 2590 (2021).Article 
ADS 

Google Scholar 
Silveira, E. M. O. et al. Forest phenoclusters for Argentina based on vegetation phenology and climate. Ecol. Appl. 32, 2526 (2022).Article 

Google Scholar 
Tatalovich, Z., Wilson, J. P. & Cockburn, M. A comparison of thiessen polygon, kriging, and spline models of potential UV exposure. Cartogr. Geogr. Inf. Sci. 33, 217–231 (2006).Article 

Google Scholar 
Choubin, B. et al. Spatiotemporal dynamics assessment of snow cover to infer snowline elevation mobility in the mountainous regions. Cold Reg. Sci. Technol. 167, 102870 (2019).Article 

Google Scholar 
Rojas, R., Flexas, J. & Coopman, R. E. Particularities of the highest elevation treeline in the world: Polylepis tarapacana Phil. as a model to study ecophysiological adaptations to extreme environments. Flora 292, 152076 (2022).Article 

Google Scholar 
Du, J. et al. Interacting effects of temperature and precipitation on climatic sensitivity of spring vegetation green-up in arid mountains of China. Agric. For. Meteorol. 269–270, 71–77 (2019).Article 
ADS 

Google Scholar 
Du, J. et al. Daily minimum temperature and precipitation control on spring phenology in arid-mountain ecosystems in China. Int. J. Climatol. 40, 2568–2579 (2020).Article 

Google Scholar 
He, Z. et al. Impacts of recent climate extremes on spring phenology in arid-mountain ecosystems in China. Agric. For. Meteorol. 260–261, 31–40 (2018).Article 
ADS 

Google Scholar 
He, Z. et al. Assessing temperature sensitivity of subalpine shrub phenology in semi-arid mountain regions of China. Agric. For. Meteorol. 213, 42–52 (2015).Article 
ADS 

Google Scholar 
Mu, C., Lu, H., Wang, B., Bao, X. & Cui, W. Short-term effects of harvesting on carbon storage of boreal Larix gmelinii–Carex schmidtii forested wetlands in Daxing’anling, northeast China. For. Ecol. Manage. 293, 140–148 (2013).Article 

Google Scholar 
Hu, T. et al. Effects of fire on soil respiration and its components in a Dahurian larch (Larix gmelinii) forest in northeast China: Implications for forest ecosystem carbon cycling. Geoderma 402, 115273 (2021).Article 
ADS 
CAS 

Google Scholar 
Nyikadzino, B., Chitakira, M. & Muchuru, S. Rainfall and runoff trend analysis in the Limpopo river basin using the Mann Kendall statistic. Phys. Chem. Earth 117, 102870 (2020).Article 

Google Scholar 
Gocic, M. & Trajkovic, S. Analysis of changes in meteorological variables using Mann-Kendall and Sen’s slope estimator statistical tests in Serbia. Glob. Planet. Change 100, 172–182 (2013).Article 
ADS 

Google Scholar 
Fang, Y. et al. Changing contribution rate of heavy rainfall to the rainy season precipitation in Northeast China and its possible causes. Atmos. Res. 197, 437–445 (2017).Article 

Google Scholar 
Piao, S. et al. Changes in satellite-derived vegetation growth trend in temperate and boreal Eurasia from 1982 to 2006. Glob. Change Biol. 17, 3228–3239 (2011).Article 
ADS 

Google Scholar 
Ahas, R., Aasa, A., Menzel, A., Fedotova, V. G. & Scheifinger, H. Changes in European spring phenology. Int. J. Climatol. 22, 1727–1738 (2002).Article 

Google Scholar 
Liang, L., Henebry, G. M., Liu, L., Zhang, X. & Hsu, L. C. Trends in land surface phenology across the conterminous United States (1982–2016) analyzed by NEON domains. Ecol. Appl. 31, e02323 (2021).Article 

Google Scholar 
Fu, Y. H. et al. Decreasing control of precipitation on grassland spring phenology in temperate China. Glob. Ecol. Biogeogr. 30, 490–499 (2020).Article 

Google Scholar 
Aze, T. Unraveling ecological signals from a global warming event of the past. Proc. Natl. Acad. Sci. U.S.A. 119, e2201495119 (2022).Article 

Google Scholar 
Menzel, A., Estrella, N. & Testka, A. Temperature response rates from long-term phenological records. Climate Res. 30, 21–28 (2005).Article 
ADS 

Google Scholar 
Wang, H., Liu, D., Lin, H., Montenegro, A. & Zhu, X. NDVI and vegetation phenology dynamics under the influence of sunshine duration on the Tibetan plateau. Int. J. Climatol. 35, 687–698 (2015).Article 

Google Scholar 
Lesica, P. & Kittelson, P. M. Precipitation and temperature are associated with advanced flowering phenology in a semi-arid grassland. J. Arid Environ. 74, 1013–1017 (2010).Article 
ADS 

Google Scholar 
Shen, M., Piao, S., Cong, N., Zhang, G. & Jassens, I. A. Precipitation impacts on vegetation spring phenology on the Tibetan Plateau. Glob. Change Biol. 21, 3647–3656 (2015).Article 
ADS 

Google Scholar 
Li, Z. et al. Spatio-temporal responses of cropland phenophases to climate change in Northeast China. J. Geog. Sci. 22, 29–45 (2012).Article 
CAS 

Google Scholar 
Badeck, F. W. et al. Responses of spring phenolgy to climate change. New Phytol. 162, 295–309 (2004).Article 

Google Scholar 
Peng, H., Xia, H., Chen, H., Zhi, P. & Xu, Z. Spatial variation characteristics of vegetation phenology and its influencing factors in the subtropical monsoon climate region of southern China. PLoS ONE 16, e0250825 (2021).Article 
CAS 

Google Scholar 
Zhang, J. et al. NIRv and SIF better estimate phenology than NDVI and EVI: Effects of spring and autumn phenology on ecosystem production of planted forests. Agric. For. Meteorol. 315, 108819 (2022).Article 
ADS 

Google Scholar 
Yu, X., Zhuang, D., Hou, X. & Chen, H. Forest phenological patterns of Northeast China inferred from MODIS data. J. Geog. Sci. 15, 239–246 (2005).Article 

Google Scholar 
Chen, X. & Xu, L. Phenological responses of Ulmus pumila (Siberian Elm) to climate change in the temperate zone of China. Int. J. Biometeorol. 56, 695–706 (2012).Article 
ADS 

Google Scholar 
Ma, X., Bai, H., He, Y. & Li, S. The vegetation RSP of Qinling Mountains based on the NDVI and the response of temperature to it. Appl. Mech. Mater. 700, 394–399 (2014).Article 

Google Scholar  More

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

Google Scholar  More

Jordan, D. S. A classification of fishes including families and genera as far as know. Stanford University Publications. Bio. Sci. 3, 79–243. https://doi.org/10.5962/bhl.title.161386 (1923).Article 

Google Scholar 
Akihito, et al. Evolutionary aspects of gobioid fishes based on an analysis of mitochondrial cytochrome b genes. Gene 259, 5–15 (2000).Article 
CAS 

Google Scholar 
Wang, H.-Y., Tsai, M.-P., Dean, J. & Lee, S.-C. Molecular phylogeny of gobioid Wshes (Perciformes: Gobioidei) based on mitochondrial 12S rRNA sequences. Mol. Phylogenet. Evol. 20, 390–408. https://doi.org/10.1016/j.ympev.2005.05.004 (2001).Article 
CAS 

Google Scholar 
Nelson, J. S., Grande, T. C. & Wilson, M. V. Fishes of the World (Wiley, 2016).Book 

Google Scholar 
Fricke, R., Eschmeyer, W. N. & Van der Laan, R. Eschmeyer’s Catalog of fishes: Genera, Species, references. (http://researcharchive.calacademy.org/research/ichthyology/catalog/fishcatmain.asp) (Accessed 15 June 2022).Guimarães-Costa, A. et al. Molecular evidence of two new species of Eleotris (Gobiiformes: Eleotridae) in the western Atlantic. Mol. Phylogenet. Evol. 98, 52–56. https://doi.org/10.1016/j.ympev.2016.01.014 (2016).Article 

Google Scholar 
Thacker, C. E. & Hardman, M. A. Molecular phylogeny of basal gobioid fishes: Rhyacichthyidae, Odontobutidae, Xenisthmidae, Eleotridae (Teleostei: Perciformes: Gobioidei). Mol. Phylogenet. Evol. 37, 858–887. https://doi.org/10.1016/j.ympev.2005.05.004 (2005).Article 
CAS 

Google Scholar 
Nordlie, F. G. Life-history characteristics of eleotrid fishes of the western hemisphere, and perils of life in a vanishing environment. Rev. Fish Biol. Fisher. 22(1), 189–224. https://doi.org/10.1007/s11160-011-9229-3 (2012).Article 

Google Scholar 
Berra, T. M. Freshwater Fish Distribution (Academic Press, 2001).
Google Scholar 
Graham, J. B. Air-Breathing Fishes: Evolution, Diversity, and Adaptation (Academic Press, 1997).Book 

Google Scholar 
Thacker, C. E. Phylogeny of Gobioidea and its placement within Acanthomorpha, with a new classification and investigation of diversification and character evolution. Copeia 1, 93–104. https://doi.org/10.1643/CI-08-004 (2009).Article 

Google Scholar 
Chakrabarty, P., Davis, M. P. & Sparks, J. S. The first record of a trans-oceanic sister-group relationship between obligate vertebrate troglobites. PLoS One 7, e44083. https://doi.org/10.1371/journal.pone.0044083 (2012).Article 
ADS 
CAS 

Google Scholar 
Agorreta, A. et al. Molecular phylogenetics of Gobioidei and phylogenetic placement of European gobies. Mol. Phylogenet. Evol. 69, 619–633. https://doi.org/10.1016/j.ympev.2013.07.017 (2013).Article 

Google Scholar 
McCraney, W. T., Thacker, C. E. & Alfaro, M. E. Supermatrix phylogeny resolves goby lineages and reveals unstable root of Gobiaria. Mol. Phylogenet. Evol. 151, 106862. https://doi.org/10.1016/j.ympev.2020.106862 (2020).Article 

Google Scholar 
Karl, S. A. & Avise, J. C. Balancing selection at allozyme loci in oysters: Implications from nuclear RFLPs. Science 256, 100. https://doi.org/10.1126/science.1348870 (1992).Article 
ADS 
CAS 

Google Scholar 
Hey, J. & Machado, C. A. The study of structured populations—New hope for a difficult and divided science. Nat. Rev. Genet. 4, 535–543. https://doi.org/10.1038/nrg1112 (2003).Article 
CAS 

Google Scholar 
Castroviejo-Fisher, S., Guayasamin, J. M., Gonzalez-Voyer, A. & Vilà, C. Neotropical diversification seen through glassfrogs. J. Biogeogr. 41, 66–80. https://doi.org/10.1111/jbi.12208 (2014).Article 

Google Scholar 
Dornburg, A., Townsend, J. P., Friedman, M. & Near, T. J. Phylogenetic informativeness reconciles ray-finned fish molecular divergence times. BMC Evol. Biol. 14, 169. https://doi.org/10.1186/s12862-014-0169-0 (2014).Article 

Google Scholar 
Hundt, P. J., Iglésias, S. P., Hoey, A. S. & Simons, A. M. A multilocus molecular phylogeny of combtooth blennies (Percomorpha: Blennioidei: Blenniidae): Multiple invasions of intertidal habitats. Mol. Phylogenet. Evol. 70, 47–56. https://doi.org/10.1016/j.ympev.2013.09.001 (2014).Article 

Google Scholar 
Olave, M., Avila, L. J., Sites, J. W. & Morando, M. Multilocus phylogeny of the widely distributed South American lizard clade Eulaemus (Liolaemini, Liolaemus). Zool. Scr. 43, 323–337. https://doi.org/10.1111/zsc.12053 (2014).Article 

Google Scholar 
Meyer, B. S., Matschiner, M. & Salzburger, W. A tribal level phylogeny of Lake Tanganyika cichlid fishes based on a genomic multi-marker approach. Mol. Phylogenet. Evol. 83, 56–71. https://doi.org/10.1016/j.ympev.2014.10.009 (2015).Article 

Google Scholar 
Jønsson, K. A. et al. A supermatrix phylogeny of corvoid passerine birds (Aves: Corvides). Mol. Phylogenet. Evol. 94, 87–94. https://doi.org/10.1016/j.ympev.2015.08.020 (2016).Article 

Google Scholar 
Li, H. & Durbin, R. Inference of human population history from individual whole-genome sequences. Nature 475(7357), 493–496. https://doi.org/10.1038/nature10231 (2011).Article 
CAS 

Google Scholar 
Frantz, R. S. X-efficiency: Theory, Evidence and Applications Vol. 2 (Springer Science & Business Media, 2013).
Google Scholar 
Bessa-Silva, A. et al. The roles of vicariance and dispersal in the differentiation of two species of the Rhinella marina species complex. Mol. Phylogenet. Evol. 145, 106723. https://doi.org/10.1016/j.ympev.2019.106723 (2020).Article 

Google Scholar 
Leutenegger, W. Maternal-fetal weight relationships in primates. Folia Primatol. 20(4), 280–293. https://doi.org/10.1159/000155580 (1973).Article 
CAS 

Google Scholar 
Yeh, J. The effect of miniaturized body size on skeletal morphology in frogs. Evolution 56(3), 628–641. https://doi.org/10.1111/j.0014-3820.2002.tb01372.x (2002).Article 

Google Scholar 
Daza, J. D. et al. An enigmatic miniaturized and attenuate whole lizard from the Mid-Cretaceous amber of Myanmar. Breviora 563(1), 1–18. https://doi.org/10.3099/MCZ49.1 (2018).Article 

Google Scholar 
Hanken, J. & Wake, D. B. Miniaturization of body size: Organismal consequences and evolutionary significance. Annu. Rev. Ecol. Evol. Syst. 24(1), 501–519. https://doi.org/10.1146/annurev.es.24.110193.002441 (1993).Article 

Google Scholar 
Britz, R. & Conway, K. W. Osteology of Paedocypris, a miniature and highly developmentally truncated fish (Teleostei: Ostariophysi: Cyprinidae). J. Morphol. 270(4), 389–412. https://doi.org/10.1002/jmor.10698 (2009).Article 
CAS 

Google Scholar 
Britz, R., Conway, K. W. & Ruber, L. Spectacular morphological novelty in a miniature cyprinid fish, Danionella dracula n. sp.. Proc. R. Soc. Lond. 276(1665), 2179–2186. https://doi.org/10.1098/rspb.2009.0141 (2009).Article 

Google Scholar 
Weitzman, S. H. & Vari, R. P. Miniaturization in South American freshwater fishes; an overview and discussion. Proc. Biol. Soc. Wash. 101(2), 444–465 (1988).
Google Scholar 
Toledo-Piza, M., Mattox, G. M. & Britz, R. Priocharax nanus, a new miniature characid from the rio Negro, Amazon basin (Ostariophysi: Characiformes), with an updated list of miniature Neotropical freshwater fishes. Neotrop. Ichthyol. 12(2), 229–246. https://doi.org/10.1590/1982-0224-20130171 (2014).Article 

Google Scholar 
Caires, R. A. & Figueiredo, J. L. Review of the genus Microphilypnus Myers, 1927 (Teleostei: Gobioidei: Eleotridae) from the lower Amazon basin, with description of one new species. Zootaxa 3036, 39–57. https://doi.org/10.11646/zootaxa.3036.1.3 (2011).Article 

Google Scholar 
Caires, R. A. Microphilypnus tapajosensis, a new species of eleotridid from the Tapajós basin, Brazil (Gobioidei: Eleotrididae). Ichthyol. Explor. Freshw. 23, 155–160 (2013).
Google Scholar 
Caires, R. A. & Guimarães-Costa, A. Family Eleotridae. In Field Guide to Amazonian Fishes (eds van Sleen, P. & Albert, J.) 388–391 (Princeton University Press, 2017).
Google Scholar 
Caires, R. A. & Toledo-Piza, M. A New species of miniature fish of the genus Microphilypnus (Gobioidei: Eleotridae) from the upper Rio Negro Basin, Amazonas Brazil. Copeia 106(1), 49–55. https://doi.org/10.1643/CI-17-634 (2018).Article 

Google Scholar 
Roberts, T.R. Leptophilypnion, a new genus with two new species of tiny central Amazonian gobioid fishes (Teleostei, Eleotridae). Aqua (2013).Gould, R. E. & Delevoryas, T. The biology of Glossopteris: Evidence from petrified seed-bearing and pollen-bearing organs. Alcheringa 1(4), 387–399 (1977).Article 

Google Scholar 
Rüber, L., Kottelat, M., Tan, H. H., Ng, P. K. & Britz, R. Evolution of miniaturization and the phylogenetic position of Paedocypris, comprising the world’s smallest vertebrate. BMC Evol. Biol. 7(1), 1–10. https://doi.org/10.1186/1471-2148-7-38 (2007).Article 
CAS 

Google Scholar 
Britz, R., Conway, K. W. & Rüber, L. Miniatures, morphology and molecules: Paedocypris and its phylogenetic position (Teleostei, Cypriniformes). Zool. J. Linn. Soc. 172(3), 556–615. https://doi.org/10.1111/zoj.12184 (2014).Article 

Google Scholar 
Bloom, D. D., Kolmann, M., Foster, K. & Watrous, H. Mode of miniaturisation influences body shape evolution in New World anchovies (Engraulidae). J. Fish Biol. 96(1), 194–201 (2019).Article 

Google Scholar 
Thacker, C. E. Molecular phylogeny of the gobioid fishes (Teleostei: Perciformes: Gobioidei). Mol. Phylogenet. Evol. 26, 354–368. https://doi.org/10.1016/S1055-7903(02)00361-5 (2003).Article 
CAS 

Google Scholar 
Birdsong, R. S., Murdy, E. O. & Pezold, F. L. A study of the vertebral column and median fin osteology in gobioid fishes with comments on gobioid relationships. Bull. Mar. Sci. 42(2), 174–214 (1988).
Google Scholar 
Thacker, C. E. Patterns of divergence in fish species separated by the Isthmus of Panama. BMC Evol. Biol. 17(1), 1–14. https://doi.org/10.1186/s12862-017-0957-4 (2017).Article 

Google Scholar 
Galván-Quesada, S. et al. Molecular phylogeny and biogeography of the amphidromous fish genus Dormitator Gill 1861 (Teleostei: Eleotridae). PLoS One 11(4), e0153538. https://doi.org/10.1371/journal.pone.0153538 (2016).Article 
CAS 

Google Scholar 
Lessios, H. A. The great American schism: Divergence of marine organisms after therise of the central American isthmus. Annu. Rev. Ecol. Evol. Syst. 2008(39), 63–92. https://doi.org/10.1146/annurev.ecolsys.38.091206.095815 (2008).Article 

Google Scholar 
Lovejoy, N. R., Albert, J. S. & Crampton, W. G. Miocene marine incursions and marine/freshwater transitions: Evidence from Neotropical fishes. J. S. Am. Earth Sci. 21, 5–13. https://doi.org/10.1016/j.jsames.2005.07.009 (2006).Article 

Google Scholar 
Cooke, G. M., Chao, N. L. & Beheregaray, L. B. Marine incursions, cryptic species and ecological diversification in Amazonia: The biogeographic history of the croaker genus Plagioscion (Sciaenidae). J. Biogeogr. 39, 724–738. https://doi.org/10.1111/j.1365-2699.2011.02635.x (2012).Article 

Google Scholar 
Bloom, D. D. & Lovejoy, N. R. On the origins of marine-derived freshwater fishes in South America. J. Biogeogr. 44(9), 1927–1938. https://doi.org/10.1111/jbi.12954 (2017).Article 

Google Scholar 
Monsch, K. A. Miocene fish faunas from the northwestern Amazonia basin (Colombia, Peru, Brazil) with evidence of marine incursions. Palaeogeogr. Palaeoclimatol. Palaeoecol. 143, 31–50. https://doi.org/10.1016/S0031-0182(98)00064-9 (1998).Article 

Google Scholar 
Hoorn, C. Marine incursions and the influence of Andean tectonics on the Miocene depositional history of northwestern Amazonia: Results of a palynostratigraphic study. Palaeogeogr. Palaeoclimatol. Palaeoecol. 105, 267–309. https://doi.org/10.1016/0031-0182(93)90087-Y (1993).Article 

Google Scholar 
Hoorn, C., Guerrero, J., Sarmiento, G. A. & Lorente, M. A. Andean tectonics as a cause for changing drainage patterns in Miocene northern South America. Geology 23, 237–240. https://doi.org/10.1130/0091-7613(1995)023%3C0237:ATAACF%3E2.3.CO;2 (1995).Article 
ADS 

Google Scholar 
Gingras, M. K., Rasanen, M. E., Pemberton, S. G. & Romero, L. P. Ichnology and sedimentology reveal depositional characteristics of bay-margin parasequences in the Miocene Amazonian foreland basin. J. Sediment. Res. 72, 871–883. https://doi.org/10.1306/052002720871 (2002).Article 
ADS 

Google Scholar 
Wesselingh, F. P. et al. Lake Pebas: A palaeoecological reconstruction of a Miocene, long-lived lake complex in western Amazonia. Cainoz. Res. 1, 35–81 (2002).
Google Scholar 
Bloom, D. D. & Lovejoy, N. R. Molecular phylogenetics reveals a pattern of biome conservatism in New World anchovies (family Engraulidae). J. Evol. Biol. 25(4), 701–715 (2012).Article 

Google Scholar 
Ward, A. B. & Azizi, E. Convergent evolution of the head retraction escape response in elongate fishes and amphibians. Zoology 107(3), 205–217. https://doi.org/10.1016/j.zool.2004.04.003 (2004).Article 

Google Scholar 
Palumbi, S. R. & Benzie, J. Large mitochondrial DNA differences between morphologically similar penaeid shrimp. Mol. Mar. Biol. Biotechnol. 1, 27–34 (1991).CAS 

Google Scholar 
Chen, W. J., Bonillo, C. & Lecointre, G. Repeatability of clades as criterion of reliability: A case study for molecular phylogeny of Acanthomorpha (Teleostei) with larger number of taxa. Mol. Phylogenet. Evol. 26, 262–288. https://doi.org/10.1016/j.gene.2008.07.016 (2003).Article 
CAS 

Google Scholar 
Chen, W. J., Miya, M., Saitoh, K. & Mayden, R. L. Phylogenetic utility of two existing and four novel nuclear gene loci in reconstructing Tree of Life of ray-finned fishes: The order Cypriniformes (Ostariophysi) as a case study. Gene 423, 125–134. https://doi.org/10.1016/j.gene.2008.07.016 (2008).Article 
CAS 

Google Scholar 
Sanger, F., Nicklen, S. & Coulson, A. R. DNA sequencing with chain-terminating inhibitors. PNAS 74(12), 5463–5467. https://doi.org/10.1073/pnas.74.12.5463 (1977).Article 
ADS 
CAS 

Google Scholar 
Edgar, R. C. MUSCLE: Multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res. 32(5), 1792–1797. https://doi.org/10.1093/nar/gkh340 (2004).Article 
CAS 

Google Scholar 
Vaidya, G., Lohman, D. J. & Meier, R. SequenceMatrix: Concatenation software for the fast assembly of multi-gene datasets with character set and codon information. Cladistics 27, 171–180 (2011).Article 

Google Scholar 
Lanfear, R., Frandsen, P. B., Wright, A. M., Senfeld, T. & Calcott, B. PartitionFinder 2: New methods for selecting partitioned models of evolution for molecular and morphological phylogenetic analyses. Mol. Biol. Evol. https://doi.org/10.1093/molbev/msw260 (2016).Article 

Google Scholar 
Heled, J. & Drummond, A. J. Bayesian inference of population size history from multiple loci. BMC Evol. Biol. 8(1), 1–15. https://doi.org/10.1186/1471-2148-8-289 (2008).Article 
CAS 

Google Scholar 
Bouckaert, R. et al. BEAST 2: A software platform for bayesian evolutionary analysis. PLoS Comput. Biol. 10(4), e1003537. https://doi.org/10.1371/journal.pcbi.1003537 (2014).Article 
CAS 

Google Scholar 
Drummond, A. J., Ho, S. Y., Phillips, M. J. & Rambaut, A. Relaxed phylogenetics and dating with confidence. PLoS Biol. 4(5), e88. https://doi.org/10.1371/journal.pbio.0040088 (2006).Article 
CAS 

Google Scholar 
Rambaut, A., Drummond, A. J., Xie, D., Baele, G. & Suchard, M. A. Posterior summarization in Bayesian phylogenetics using Tracer 1.7. Syst. Biol. 67(5), 901. https://doi.org/10.1093/sysbio/syy032 (2018).Article 
CAS 

Google Scholar 
Drummond, A. J. & Rambaut, A. BEAST: Bayesian evolutionary analysis by sampling trees. BMC Evol. Biol. 7, 214. https://doi.org/10.1186/1471-2148-7-214 (2007).Article 
CAS 

Google Scholar 
Rambaut, A. FigTree, a graphical viewer of phylogenetic trees (Version 1.4.3) (2017).Betancur-R, R. et al. Phylogenetic classification of bony fishes. BMC Evol. Biol. 17(1), 1–40. https://doi.org/10.1186/s12862-017-0958-3 (2017).Article 

Google Scholar 
Jones, G. Algorithmic improvements to species delimitation and phylogeny estimation under the multispecies coalescent. J. Math. Biol. 74, 447–467 (2017).Article 
MathSciNet 
MATH 

Google Scholar  More

FAO. Global forest resources assessment. www.fao.org/publications (2015).Finlayson, M. et al. A Report of the Millennium Ecosystem Assessment. (The Cropper Foundation, 2005).Lal, R., & Lorenz, K. In Recarbonization of the Biosphere: Ecosystems and the Global Carbon Cycle (eds Lal, R., Lorenz, K., Hüttl, R. F., Schneider, B. U. & von Braun, J.) Ch. 9 (Springer, 2012).Gilliam, F. S. Forest ecosystems of temperate climatic regions: from ancient use to climate change. N. Phytologist 212, 871–887 (2016).Article 

Google Scholar 
de Gouvenain, R. C. & Silander, J. A. Temperate forests in Reference Module in Life Sciences (Elsevier, 2017).Keith, S. A., Newton, A. C., Morecroft, M. D., Bealey, C. E. & Bullock, J. M. Taxonomic homogenization of woodland plant communities over 70 years. Proc. R. Soc. B: Biol. Sci. 276, 3539–3544 (2009).Article 

Google Scholar 
Rackham, O. Ancient woodlands: modern threats. N. Phytologist 180, 571–586 (2008).Article 

Google Scholar 
Bernhardt-Römermann, M. et al. Drivers of temporal changes in temperate forest plant diversity vary across spatial scales. Glob. Chang. Biol. 21, 3726–3737 (2015).Article 
ADS 

Google Scholar 
Waller, D. M. & Alverson, W. S. The white-tailed deer: a keystone herbivore. Wildl. Soc. Bull. 25, 217–226 (1997).
Google Scholar 
Ramirez, J. I. Uncovering the different scales in deer–forest interactions. Ecol. Evol. 11, 5017–5024 (2021).Article 

Google Scholar 
Rooney, T. P., Wiegmann, S. M., Rogers, D. A. & Waller, D. M. Biotic impoverishment and homogenization in unfragmented forest understory communities. Conserv. Biol. 18, 787–798 (2004).Stockton, S. A., Allombert, S., Gaston, A. J. & Martin, J. L. A natural experiment on the effects of high deer densities on the native flora of coastal temperate rain forests. Biol. Conserv 126, 118–128 (2005).Article 

Google Scholar 
Hegland, S. J., Lilleeng, M. S. & Moe, S. R. Old-growth forest floor richness increases with red deer herbivory intensity. Ecol. Manag. 310, 267–274 (2013).Article 

Google Scholar 
Simončič, T., Bončina, A., Jarni, K. & Klopčič, M. Assessment of the long-term impact of deer on understory vegetation in mixed temperate forests. J. Veg. Sci. 30, 108–120 (2019).Article 

Google Scholar 
Vild, O. et al. The paradox of long-term ungulate impact: increase of plant species richness in a temperate forest. Appl. Veg. Sci. 20, 282–292 (2017).Article 

Google Scholar 
Russell, F. L., Zippin, D. B. & Fowler, N. L. Effects of white-tailed deer (Odocoileus virginianus) on plants, plant populations and communities: a review. Am. Midl. Nat. 146, 1–26 (2001).Article 

Google Scholar 
Öllerer, K. et al. Beyond the obvious impact of domestic livestock grazing on temperate forest vegetation–A global review. Biol. Conserv. 237, 209–219 (2019).Article 

Google Scholar 
Borer, E. T. et al. Nutrients cause grassland biomass to outpace herbivory. Nat. Commun. 11, 1–8 (2020).Article 
ADS 

Google Scholar 
Kaarlejärvi, E., Eskelinen, A. & Olofsson, J. Herbivores rescue diversity in warming tundra by modulating trait-dependent species losses and gains. Nat. Commun. 8, 1–8 (2017).
Google Scholar 
Simkin, S. M. et al. Conditional vulnerability of plant diversity to atmospheric nitrogen deposition across the United States. Proc. Natl Acad. Sci. USA 113, 4086–4091 (2016).Article 
ADS 
CAS 

Google Scholar 
Bobbink, R. et al. Global assessment of nitrogen deposition effects on terrestrial plant diversity: A synthesis. Ecol. Appl. 20, 30–59 (2010).Article 
CAS 

Google Scholar 
Reinecke, J., Klemm, G. & Heinken, T. Vegetation change and homogenization of species composition in temperate nutrient deficient Scots pine forests after 45 yr. J. Veg. Sci. 25, 113–121 (2014).Article 

Google Scholar 
Speed, J. D. M., Austrheim, G., Kolstad, A. L. & Solberg, E. J. Long-term changes in northern large-herbivore communities reveal differential rewilding rates in space and time. PLoS ONE 14, e0217166 (2019).Article 
CAS 

Google Scholar 
Valente, A. M., Acevedo, P., Figueiredo, A. M., Fonseca, C. & Torres, R. T. Overabundant wild ungulate populations in Europe: management with consideration of socio-ecological consequences. Mamm. Rev. 50, 353–366 (2020).Article 

Google Scholar 
Linnell, J. D. C. et al. The challenges and opportunities of coexisting with wild ungulates in the human-dominated landscapes of Europe’s Anthropocene. Biol. Conserv. 244, 108500 (2020).Waller, D. M. The Herbaceous Layer in Forests of Eastern North America (ed. Gilliam, F.) Ch. 16 (Oxford Univ. Press, 2014).Kerley, G. I. H., Kowalczyk, R. & Cromsigt, J. P. G. M. Conservation implications of the refugee species concept and the European bison: king of the forest or refugee in a marginal habitat? Ecography 35, 519–529 (2011).Svenning, J. C. A review of natural vegetation openness in north-western Europe. Biol. Conserv 104, 133–148 (2002).Article 

Google Scholar 
Sandom, C. J., Ejrnaes, R., Hansen, M. D. D. & Svenning, J. C. High herbivore density associated with vegetation diversity in interglacial ecosystems. Proc. Natl Acad. Sci. USA 111, 4162–4167 (2014).Article 
ADS 
CAS 

Google Scholar 
Ramirez, J. I., Jansen, P. A., den Ouden, J., Goudzwaard, L. & Poorter, L. Long-term effects of wild ungulates on the structure, composition and succession of temperate forests. Ecol. Manag. 432, 478–488 (2019).Article 

Google Scholar 
Ramirez, J. I., Jansen, P. A. & Poorter, L. Effects of wild ungulates on the regeneration, structure and functioning of temperate forests: A semi-quantitative review. Ecol. Manag. 424, 406–419 (2018).Article 

Google Scholar 
Albert, A. et al. Seed dispersal by ungulates as an ecological filter: a trait-based meta-analysis. Oikos 124, 1109–1120 (2015).Article 

Google Scholar 
McNaughton, S. J. Grazing lawns: on domesticated and wild grazers. Am. Nat. 128, 937–939 (1986).Article 

Google Scholar 
Cromsigt, J. P. G. M. & Kuijper, D. P. J. Revisiting the browsing lawn concept: evolutionary Interactions or pruning herbivores? Perspect. Plant Ecol. 13, 207–215 (2011).Article 

Google Scholar 
Ramirez, J. I. et al. Temperate forests respond in a non-linear way to a population gradient of wild deer. Forestry 94, 502–511 (2021).Article 

Google Scholar 
Boulanger, V. et al. Ungulates increase forest plant species richness to the benefit of non‐forest specialists. Glob. Chang. Biol. 24, e485–e495 (2018).Article 

Google Scholar 
Kirby, K. J. The impact of deer on the ground flora of British broadleaved woodland. Forestry 74, 219–229 (2001).Article 

Google Scholar 
Royo, A. A., Collins, R., Adams, M. B., Kirschbaum, C. & Carson, W. P. Pervasive interactions between ungulate browsers and disturbance regimes promote temperate forest herbaceous diversity. Ecology 91, 93–105 (2010).Happonen, K. et al. Trait-based responses to land use and canopy dynamics modify long-term diversity changes in forest understories. Glob. Ecol. Biogeogr. 30, 1863–1875 (2021).Article 

Google Scholar 
Peñuelas, J. & Sardans, J. The global nitrogen-phosphorus imbalance. Science 375, 266–267 (2022).Article 
ADS 

Google Scholar 
Staude, I. R. et al. Replacements of small- by large-ranged species scale up to diversity loss in Europe’s temperate forest biome. Nat. Ecol. Evol. 4, 802–808 (2020).Article 

Google Scholar 
Newbold, T. et al. Widespread winners and narrow-ranged losers: Land use homogenizes biodiversity in local assemblages worldwide. PLoS Biol. 16, e2006841 (2018).Article 

Google Scholar 
Verheyen, K. et al. Driving factors behind the eutrophication signal in understorey plant communities of deciduous temperate forests. Br. Ecol. Soc. J. Ecol. 100, 352–365 (2012).
Google Scholar 
Gilliam, F. S. Response of the herbaceous layer of forest ecosystems to excess nitrogen deposition. J. Ecol. 94, 1176–1191 (2006).Article 
CAS 

Google Scholar 
de Schrijver, A. et al. Cumulative nitrogen input drives species loss in terrestrial ecosystems. Glob. Ecol. Biogeogr. 652, 803–816 (2011).Article 

Google Scholar 
de Frenne, P. et al. Light accelerates plant responses to warming. Nat. Plants 1, 15110 (2015).Article 

Google Scholar 
Baeten, L. et al. Herb layer changes (1954-2000) related to the conversion of coppice-with-standards forest and soil acidification. Appl. Veg. Sci. 12, 187–197 (2009).Article 

Google Scholar 
Becker, T., Spanka, J., Schröder, L. & Leuschner, C. Forty years of vegetation change in former coppice-with-standards woodlands as a result of management change and N deposition. Appl. Veg. Sci. 20, 304–313 (2017).Article 

Google Scholar 
van Calster, H. et al. Diverging effects of overstorey conversion scenarios on the understorey vegetation in a former coppice-with-standards forest. Ecol. Manag. 256, 519–528 (2008).Article 

Google Scholar 
Luyssaert, S. et al. The European carbon balance. Part 3: forests. Glob. Chang. Biol. 16, 1429–1450 (2010).Article 
ADS 

Google Scholar 
Kirby, K. J. et al. Five decades of ground flora changes in a temperate forest: the good, the bad and the ambiguous in biodiversity terms. Ecol. Manag. 505, 119896 (2022).Article 

Google Scholar 
Hautier, Y., Niklaus, P. A. & Hector, A. Competition for light causes plant biodiversity loss after eutrophication. Science 324, 636–638 (2009).Article 
ADS 
CAS 

Google Scholar 
Kowalczyk, R., Kamiński, T. & Borowik, T. Do large herbivores maintain open habitats in temperate forests? For. Ecol. Manag. 494, 119310 (2021).Dormann, C. F. et al. Plant species richness increases with light availability, but not variability, in temperate forests understorey. BMC Ecol. 20, 1–9 (2020).Article 

Google Scholar 
Dirnböck, T. et al. Forest floor vegetation response to nitrogen deposition in Europe. Glob. Chang. Biol. 20, 429–440 (2014).Article 
ADS 

Google Scholar 
Perring, M. P. et al. Understanding context dependency in the response of forest understorey plant communities to nitrogen deposition. Environ. Pollut. 242, 1787–1799 (2018).Article 
CAS 

Google Scholar 
Anderson, T. M. et al. Herbivory and eutrophication mediate grassland plant nutrient responses across a global climatic gradient. Ecology 99, 822–831 (2018).Article 

Google Scholar 
Gough, L. & Grace, J. B. Herbivore effects on plant species density at varying productivity levels. Ecology 79, 1586–1594 (1998).Article 

Google Scholar 
Eskelinen, A., Harpole, W. S., Jessen, M.-T., Virtanen, R. & Hautier, Y. Light competition drives herbivore and nutrient effects on plant diversity. Nature 611, 301–305 (2022).Knight, T. M., Dunn, J. L., Smith, L. A., Davis, J. A. & Kalisz, S. Deer facilitate invasive plant success in a Pennsylvania forest understory. Nat. Areas 29, 110–116 (2009).Article 

Google Scholar 
Beguin, J., Pothier, D. & Côté, S. D. Deer browsing and soil disturbance induce cascading effects on plant communities: a multilevel path analysis. Ecol. Appl. 21, 439–451 (2011).Gilliam, F. S. et al. Twenty-five-year response of the herbaceous layer of a temperate hardwood forest to elevated nitrogen deposition. Ecosphere 7, e01250 (2016).Article 

Google Scholar 
de Frenne, P. et al. Microclimate moderates plant responses to macroclimate warming. Proc. Natl Acad. Sci. USA 110, 18561–18565 (2013).Article 
ADS 

Google Scholar 
Hedwall, P. O. et al. Half a century of multiple anthropogenic stressors has altered northern forest understory plant communities. Ecol. Appl. 29, e01874 (2019).Perring, M. P. et al. Global environmental change effects on plant community composition trajectories depend upon management legacies. Glob. Chang. Biol. 24, 1722–1740 (2018).Article 
ADS 

Google Scholar 
Boulanger, V. et al. Decreasing deer browsing pressure influenced understory vegetation dynamics over 30 years. Ann. Sci. 72, 367–378 (2015).Article 

Google Scholar 
Bernes, C. et al. Manipulating ungulate herbivory in temperate and boreal forests: effects on vegetation and invertebrates. A systematic review. Environ. Evid. 7, 1–32 (2018).Article 

Google Scholar 
Reimoser, F. Steering the impacts of ungulates on temperate forests. J. Nat. Conserv. 10, 243–252 (2003).Article 

Google Scholar 
Vavra, M., Parks, C. G. & Wisdom, M. J. Biodiversity, exotic plant species, and herbivory: the good, the bad, and the ungulate. Ecol. Manag. 246, 66–72 (2007).Article 

Google Scholar 
Depauw, L. et al. Light availability and land-use history drive biodiversity and functional changes in forest herb layer communities. J. Ecol. 108, 1411–1425 (2020).Article 
CAS 

Google Scholar 
Chevaux, L. et al. Effects of stand structure and ungulates on understory vegetation in managed and unmanaged forests. Ecol. Appl. 32, e01874 (2022).Gordon, I. J. Browsing and grazing ruminants: are they different beasts? Ecol. Manag. 181, 13–21 (2003).Article 

Google Scholar 
Brasseur, B. et al. What deep‐soil profiles can teach us on deep‐time pH dynamics after land use change? Land Degrad. Dev. 29, 2951–2961 (2018).Article 

Google Scholar 
Schmitz, A. et al. Responses of forest ecosystems in Europe to decreasing nitrogen deposition. Environ. Pollut. 244, 980–994 (2019).Article 
CAS 

Google Scholar 
Dirnböck, T. et al. Currently legislated decreases in nitrogen deposition will yield only limited plant species recovery in European forests. Environ. Res. Lett. 13, 125010 (2018).Article 

Google Scholar 
Peterken, G. F. Natural Woodland: Ecology and Conservation in Northern Temperate Regions (Cambridge Univ. Press, 1996).Chamberlain, S. A. & Boettiger, C. R Python, and Ruby clients for GBIF species occurrence data. preprint. PeerJ Preprints 5, e3304v1 (2017).Chamberlain, S. A. & Szöcs, E. taxize: taxonomic search and retrieval in R. F1000Res 2, 191 (2013).Article 

Google Scholar 
Hédl, R., Kopecký, M. & Komárek, J. Half a century of succession in a temperate oakwood: from species-rich community to mesic forest. Divers Distrib. 16, 267–276 (2010).Article 

Google Scholar 
Giménez-Anaya, A., Herrero, J., Rosell, C., Couto, S. & García-Serrano, A. Food habits of wild boars (Sus scrofa) in a Mediterranean coastal wetland. Wetlands 28, 197–203 (2008).Article 

Google Scholar 
Barrios-Garcia, M. N. & Ballari, S. A. Impact of wild boar (Sus scrofa) in its introduced and native range: a review. Biol. Invasions 14, 2283–2300 (2012).Article 

Google Scholar 
Andersen, R. et al. An overview of the progress and challenges of peatland restoration in Western Europe. Restor. Ecol. 25, 271–282 (2017).Article 

Google Scholar 
Faurby, S. et al. PHYLACINE 1.2: the phylogenetic atlas of mammal macroecology. Ecology 99, 2626 (2018).Article 

Google Scholar 
van den Berg, L. J. L. et al. Evidence for differential effects of reduced and oxidised nitrogen deposition on vegetation independent of nitrogen load. Environ. Pollut. 208, 890–897 (2016).Article 

Google Scholar 
McNaughton, S. J., Oesterheld, M., Frank, D. A. & Williams, K. J. Ecosystem-level patterns of primary productivity and herbivory in terrestrial habitats. Nature 341, 142–144 (1989).Article 
ADS 
CAS 

Google Scholar 
Koerner, S. E. et al. Change in dominance determines herbivore effects on plant biodiversity. Nat. Ecol. Evol. 2, 1925–1932 (2018).Article 

Google Scholar 
Fréjaville, T. & Garzón, M. B. The EuMedClim database: yearly climate data (1901-2014) of 1 km resolution grids for Europe and the Mediterranean Basin. Front. Ecol. Evol. 6, 1–5 (2018).Article 

Google Scholar 
Al‐Yaari, A. et al. Asymmetric responses of ecosystem productivity to rainfall anomalies vary inversely with mean annual rainfall over the conterminous United States. Glob. Chang. Biol. 26, 6959–6973 (2020).Article 
ADS 

Google Scholar 
Szabó, P. & Hédl, R. Advancing the integration of history and ecology for conservation. Conserv. Biol. 25, 680–687 (2011).Article 

Google Scholar 
Hedges, L. V., Gurevitch, J. & Curtis, P. S. The meta-analysis of response ratios in experimental ecology. Spec. Feature Ecol. 80, 1150–1156 (1999).
Google Scholar 
Hillebrand, H. et al. Biodiversity change is uncoupled from species richness trends: consequences for conservation and monitoring. J. Appl. Ecol. 55, 169–184 (2018).Article 

Google Scholar 
Holz, H., Segar, J., Valdez, J. & Staude, I. R. Assessing extinction risk across the geographic ranges of plant species in Europe. Plants People Planet 4, 303–311 (2022).Article 

Google Scholar 
Staude, I. R. et al. Directional turnover towards larger‐ranged plants over time and across habitats. Ecol. Lett. 25, 466–482 (2021).Article 

Google Scholar 
Ellenberg, H., Weber, H. E., Düll, R., Wirth, V. & Werner, W. Zeigerwerte von Pflanzen in Mitteleuropa (Verlag Wrich Goltze, 2001).Chytrý, M., Tichý, L., Dřevojan, P., Sádlo, J. & Zelený, D. Ellenbergtype indicator values for the Czech flora. Preslia 90, 83–103 (2018).Bürkner, P.-C. brms: an R package for Bayesian multilevel models using Stan. J. Stat. Softw. 80, 1–28 (2017).Article 

Google Scholar 
Brooks, S. P. & Gelman, A. General methods for monitoring convergence of iterative simulations. J. Comput. Graph. Stat. 7, 434–455 (1998).MathSciNet 

Google Scholar 
Dushoff, J., Kain, M. P. & Bolker, B. M. I can see clearly now: reinterpreting statistical significance. Methods Ecol. Evol. 10, 756–759 (2019).Article 

Google Scholar 
Bradshaw, L. & Waller, D. M. Impacts of white-tailed deer on regional patterns of forest tree recruitment. Ecol. Manag. 375, 1–11 (2016).Article 

Google Scholar 
McGarvey, J. C., Bourg, N. A., Thompson, J. R., McShea, W. J. & Shen, X. Effects of twenty years of deer exclusion on woody vegetation at three life-history stages in a mid-atlantic temperate deciduous forest. Northeast. Nat. 20, 451–468 (2013).Nuttle, T., Ristau, T. E. & Royo, A. A. Long-term biological legacies of herbivore density in a landscape-scale experiment: forest understoreys reflect past deer density treatments for at least 20 years. J. Ecol. 102, 221–228 (2013). More

Correction to: Communications Biology https://doi.org/10.1038/s42003-020-1022-1, published online 11 June 2020.In the original version of the Perspective, a unit conversion error affected calculations for cereal rye, triticale, barley, and oats. Further, berseem clover yield estimates were mistranscribed from the original source. These mistakes led to errors in Supplementary Data 1, Figure 2 and in the presentation of the data in the text.Supplementary Data 1 has now been replaced with a file containing the correct numbers.Figure 2 has been corrected:Original figure 2New figure 2The Abstract stated: “In this Perspective, we estimate land use requirements to supply the United States maize production area with cover crop seed, finding that across 18 cover crops, on average 3.8% (median 2.0%) of current production area would be required, with the popular cover crops rye and hairy vetch requiring as much as 4.5% and 11.9%, respectively”.The text should read: “In this Perspective, we estimate land use requirements to supply the United States maize production area with cover crop seed, finding that across 18 cover crops, on average 2.4% (median 2.1%) of current production area would be required, with the popular cover crops rye and hairy vetch requiring as much as 4.8% and 11.9%, respectively”.In the 1st paragraph of the right hand column on page 2, the text said: “(…), we find that the land requirements for production of cover crop seed would be on average 1.4 million hectares (median 746,000 ha), which is equivalent to 3.8% (median 2.0%) of the U.S. maize farmland. Rye (Secale cereale L.) – a midrange seed yielding cover crop and one of the most commonly used in the corn belt, would require as much as 1,661,000 hectares (4.5% of maize farmland), (…)”The text should read: “(…) we find that the land requirements for production of cover crop seed would be on average 892,526 hectares (median 774,417 ha), which is equivalent to 2.4% (median 2.1%) of the U.S. maize farmland. Rye (Secale cereale L.) – a midrange seed yielding cover crop and one of the most commonly used in the corn belt, would require as much as 1,779,770 hectares (4.8% of maize farmland), (…)”On page 3, second paragraph the text said: “Cover cropping the entire U.S. maize area would require the equivalent of as much as 18% (rye) to 49% (hairy vetch) (…)”The text should read: “Cover cropping the entire U.S. maize area would require the equivalent of as much as 19% (rye) to 49% (hairy vetch) (…)”This errors have now been corrected in the Perspective Article. More