Ecology

Candidate bird deterrentsWe identified four candidate bird deterrents that were suitable for deployment within a SN-toxic baiting program (Fig. 2). Specifically, we searched published studies and vendor websites to identify candidate bird deterrents that had a proven record of deterring birds, or features that we expected would deter all birds after a deployment of SN-toxic bait while not deterring wild pigs. These features included: (1) not deterring wild pigs (i.e., user programmable operating hours for after wild pigs visits or being bird-specific), (2) aversive to birds (i.e., erratic movements or irritating to birds), and (3) remotely operated (i.e., battery operated or effects lasting ~ 12 h if user applied).Figure 2Examples of potential bird deterrents tested in in north-central Colorado, USA during April–May 2020, including (A) control, no deterrent, (B) 7.5% concentration of methyl anthranilate, (C) a metal grate, (D), an inflatable scarecrow, and (E) a scare dancer. Photos property of USDA.Full size imageWe selected two frightening devices that offered visual and auditory stimuli, were battery-powered, and programmable to have a user-specified start time. The first frightening device was a 1.8 m inflatable scare dancer (Snake 6 ft Cordless Inflatable Scarecrow, AirCrow LLC, Lake Charles, LA, USA). The scare dancer was a yellow nylon tube shaped like a snake and inflated by a small fan and control unit powered by a 12 V battery connected to a programmable control panel. If using the scare dancer for SN-toxic bait deployment, our strategy would be to program the device to operate continuously starting 1 h before first light the morning after toxic bait was deployed. Our expectation would be that wild pigs would have already visited bait sites and consumed SN-toxic bait prior to scare dancer activation. Once activated, the scare dancer would deter non-targets away from any spilled SN-toxic bait during the morning after toxic baiting until operators arrived to clean the site.The second frightening device was an inflatable scarecrow called the Scarey Man Birdscarer (Clarratts Ltc, United Kingdom). This device was also powered by a small fan using a 12 V battery, activated by a timer, and inflated for 25 s every 18 min accompanied by an audible 112 db siren. The timing of the inflation could not be altered. The blaze-orange inflatable scarecrow bobbed up and down as it inflated and deflated, and emitted a siren wail. Our strategy with the inflatable scarecrow, following SN-toxic bait deployment, would be the same as the scare dancer, except the inflatable scarecrow could not be programmed to operate continuously.For the physical barrier treatment, we constructed a metal grate using a 2.4 m × 1.2 m sheet of #13-gauge steel diamond-shaped expanded metal. The maximum openings of the expanded metal were 1.0 cm and were raised (i.e., tapered upwards) to facilitate bait falling through the grate. We constructed the grate to sit 9.0 cm above ground using a frame of standard construction lumber. We also tapered the top of the wooden frame to reduce surface area and facilitate bait falling through the grate. If using the grate for SN-toxic bait deployment, our strategy would be to put the bait station on top of the grate. Our expectation would be that wild pigs would stand on the grate to access the bait station, and spilled particles would fall under the grate and be inaccessible to non-target animals.The chemical repellent treatment we tested was Avian Migrate™ Goose and Bird Repellent (Avian Enterprises, Jupiter, FL, USA) which contained 14.5% methyl anthranilate. Avian Migrate required dilution with water for all applications. We followed the label instructions for spot repelling, and used the strongest dilution recommended at 50:50 Avian Migrate and water, resulting in 7.5% methyl anthranilate. We used a hand-pump-pressurized garden sprayer to apply 500 ml of the mixture to a 3 × 3 m area which resulted in an even and thorough coating of the area. Aversion to methyl anthranilate may be a learned behavior as an irritant for birds36, therefore would need to be applied daily for 1–2 days prior to SN-toxic bating. If using the repellent for SN-toxic bait deployment, our strategy would be to spray the ground immediately surrounding bait stations for 2 nights prior to deploying toxic bait, and the night of toxic baiting. Our expectation would be that by the 3rd night of application non-target birds would be repelled from consuming particles of spilled bait that fell on the treated ground; after which, we could safely deploy SN-toxic bait.Field study on deterrent effectiveness for birdsWe initially selected and pre-baited ~ 60 sites in north-central CO using 5 kg of bird seed (Deluxe Blend Bird Seed, Wild Birds Unlimited, Fort Collins, CO, USA). Sites were selected in diverse land covers that were likely to hold small passerine birds, such as thickets, wind rows, near water sources, or along shelter belts; and based on distance to nearby sites (i.e., goal of  > 500 m to nearest site). We cleared sites of tall grass and debris to ease discovery and access to the bird seed by smaller birds. We visited sites every 2–3 days to replenish and maintain ~ 2 kg of bait at the sites. We pre-baited sites for ~ 4 weeks to ensure birds were well-acclimated to visiting sites daily.We monitored visitation to sites using remote cameras (RECONYX PC900, RECONYX Inc, Holmen, WI, USA) mounted on T-post approximately 5 m from the bait pile, 1.5 m above ground, and angled down at 70° to provide a consistent field of view at each site. Cameras were programmed to record time-lapse imagery every 2 min (i.e., 720 images/day) which was used to calculate indices of species visitation. We used the Colorado Parks and Wildlife Photo Database to process all time-lapse imagery (Ivan and Newkirk 2016). For each image, a single observer recorded presence and count of each unique species present. We selected the best 20 sites (Fig. 1) based on the greatest rates of bird visitation, greatest diversity of bird species visiting, and lowest presence of other species that consumed large quantities of the bird seed (e.g., raccoons, deer, skunks).For the trial, we randomly assigned a deterrent treatment (i.e., inflatable scarecrow, metal grate, methyl anthranilate) or control (i.e., no deterrent method) to five sites each. We re-used the control sites to test the scare dancer after testing the initial four treatments, because the scare dancers were received later than first three treatments. We visited bait sites daily and weighed the amount of bird seed remaining to calculate the amount consumed with digital scales (MeasureTek GGS_42964, MeasureTek Scale Co, Ltd, Vancouver, BC, Canada). We replenished each site to ensure ~ 2 kg of fresh bird seed was available each day.The trials were seven consecutive days (Table 1). We focused on species visitation from 1 h before first light (~ 0500 h) to midday (1200 h) each day, because this time period represented the critical hours in which hazards occurred at toxic bait sites22,24. We visited the bait sites between 1200 and 1400 h each day to replenish bait and prepare sites for the following day. The 7-day trial consisted of:

Days 1–2 = Pre-baiting days. No deterrent deployed.

Day 3 = Acclimation day. We deployed the deterrent devices but did not activate. Scare dancers were installed on a t-post 1.5 m above the bait sites. Inflatable scarecrows were placed on the ground 3 m away from the bait sites. Metal grates were deployed 3 m away from the bait sites. Methyl anthranilate was sprayed for first time in the 3 × 3 m area surrounding bait sites to initiate the learned repellency.

Day 4 = Pre-treatment day. This was the day we collected pre-treatment data (i.e., consumption and remote camera data) for comparison with treatment and post-treatment below. All deterrent devices remained inactive as described for acclimation day. The methyl anthranilate was sprayed in the same manner as before for the second time.

Day 5 = Treatment day. Both frightening devices were activated at 1 h prior to first light. The metal grate was installed over the bird seed. Methyl anthranilate was sprayed in the same manner as before for the third and final time.

Day 6 = Post-treatment day. All deterrent devices were inactivated but left in place similar to the pre-treatment day. The metal grate was moved 3 m away from the bait site. No methyl anthranilate was sprayed.

Day 7 = Removal day. We removed all our cameras and deterrent devices and ceased re-baiting at all sites.

Table 1 Strategies used to evaluate effectiveness of bird deterrents during a 7-day trial in north-central Colorado, USA during April–May 2020.Full size tableFor each site, we calculated an index of the number of passerine birds observed in each two-min time-lapse image (rate = average number of birds/two mins) during morning hours (i.e., 0500–1200) for the morning of pre-treatment, treatment, and post-treatment. We compared indices among each of the 3 days and five treatments using negative binomial mixed models and log-links with package glmmTMB37 in Program R v3.6.338. We used offsets of the number of hours monitored and site ID as a random effect to account for repeated (i.e., daily) measures taken at each site. We did not analyze for other species (i.e., predatory birds and mammals) because visitations were rare. For all analyses we calculated and examined the 95% confidence intervals (CIs) surrounding the regression coefficients (β) for non-overlap of zero to indicate statistical and biological differences.Effects of deterrents on captive wild pigsWe evaluated whether the deterrents influenced feeding behaviors of captive wild pigs. Specifically, we evaluated how wild pigs responded to the metal grate and methyl anthranilate, because these deterrent strategies would need to be in place as wild pigs visited bait sites, and we wanted to ensure wild pigs would not be deterred from feeding. Contrarily, neither of the deterrent devices should be encountered by wild pigs because these devices would be operated on a timer and set to activate after wild pigs visited toxic baiting sites. Therefore, we did not evaluate those treatments with captive wild pigs.For testing methyl anthranilate, we randomly selected and placed three captive wild pigs from the larger holding pen (i.e., two males and one female) into three 0.02 ha pens, respectively. We replicated this design twice, for a total of six pens (n = 18 wild pigs) tested. The wild pigs in each pen were acclimated for one night to the new pens and to feeding from two identical feed troughs (1.8 × 0.3 × 0.1 m) placed 3.2 m apart. Each night we fed ~ 10 kg of whole kernel corn in each trough and weighed any remaining corn the following morning. A 2-choice feeding test was conducted on nights two, three, and four, where we applied methyl anthranilate to a 3 × 3 m area surrounding one of the troughs using the same mixture as described above in CO. For the other trough, we did not apply methyl anthranilate to the surrounding soil. We applied the methyl anthranilate and whole kernel corn each evening of the 3-day treatment period.For testing the metal grate, we randomly selected and placed four captive wild pigs from the larger holding pen into two 0.2 ha pens, respectively. We replicated this design twice, for a total of four pens (n = 16 wild pigs) tested. A single feed trough (1.8 × 0.3 × 0.1 m) was placed in each pen. We placed the metal grate under the trough in one pen where it remained for the three nights of study. Two kg of pelleted sow ration were fed in each pen on night 1. On night two, ~ 10 kg of a placebo SN-toxic bait (i.e., HOGGONE without SN) and 1 kg of pelleted sow ration were fed in each pen. On night three we offered just 10 kg of placebo bait to evaluate whether spilled particles of the peanut paste-based bait16 would stick to the metal grate. We ceased testing the metal grate after the second replicate because we observed that wild pigs spilled small particles of the placebo bait which stuck to the top of the metal grate in the first replicate, followed by 100% aversion by wild pigs to the metal grate in the second replicate, rendering the metal grate a non-viable option for operational use.For the methyl anthranilate, we compared proportions of whole-kernel corn consumed in the 2-choice test using a linear model in Program R. We evaluated the interaction of treatment × night to determine if the application of methyl anthranilate influenced the amount of corn wild pigs consumed over time. We also tested the reduced model without the interaction to best interpret the unconditional main effects39. We did not analyze data from the metal grate treatment because the evaluation was stopped early, and the results were clear.Field evaluation of deterrent with toxic baitFor the final phase of this study, we evaluated the most effective deterrent identified in the first phase of the study (i.e., scare dancer deterrent device, see results) and implemented this deterrent device into a SN-toxic toxic baiting program for wild pigs in north-central TX. We followed methodologies established in previous studies (Table 2) to initiate a SN-baiting program24,40,41,42. Specifically, we initially deployed ~ 30 bait sites by placing ~ 11 kg of whole-kernel corn on the ground at locations with recent sign of wild pigs (e.g., fresh tracks, feces, wallowing, rooting). We installed one remote camera on a t-post 5 m away from each bait site, 1.5 m above ground, and angled down at 70°. We programmed cameras to capture time-lapse images every 5 min (i.e., 288 images/day). We revisited bait sites every day for 5 days to refresh bait (i.e., maintain 11 kg of corn) and view camera images for wild pigs. After day 5, we selected the 10 best sites (Fig. 1) using the highest ranked sites from this ranking system: (1) consistent wild pig visitation (i.e., ≥ 2 days in a row), (2) consistent visitation by a family group of wild pigs (i.e., ≥ 1 female with multiple juveniles or piglets), (3) consistent visitation by multiple family groups (4) consistent visitation of independent family groups not visiting other sites42. We also made sure to select bait sites that were  > 500 m apart to maintain independence among the groups of pigs visiting each site41,43.Table 2 Baiting strategy to locate, pre-bait, and train wild pigs to use bait stations and consume SN-toxic bait used in north-central Texas, USA during July 2020.Full size tableWe deployed wild pig-specific bait stations20 with ~ 13 kg of magnetic resistance on the lids21 at the 10 final sites and initiated a series of conditioning phases to acclimate wild pigs to open and consume bait from inside the bait stations (Table 2). We deployed two bait stations at sites with ≥ 10 wild pigs to ensure all wild pigs had sufficient access to bait. We deployed bait stations 10–30 m away from initial pre-baiting sites (where we originally placed corn on the ground) to reduce visitation by non-target animals that may be attracted to residual particles of corn. Where cattle were present, we also constructed 3-strand barbed-wire fences around the site to exclude them from accessing SN-toxic bait.We randomly selected five sites to deploy the deterrent devices, and five sites as controls (no deterrent devices). Three days prior to deploying SN-toxic bait, we deployed the deterrent devices but left them inactive to condition wild pigs to the presence of the devices. We mounted the deterrent devices on T-posts approximately 1.8 m above ground directly over each bait station with the battery box secured at the base of the T-post (Fig. 3). When we deployed SN-toxic bait, we programmed the deterrent devices to activate at 0520 h the next morning (i.e., 1 h before first-light). We waited until 0900–1200 h the next morning before visiting bait sites to allow ample testing time of the deterrent devices to deter birds, and to simulate realistic use in an operational setting. When we arrived at the bait site, we deactivated the deterrent devices and cleaned the surrounding area of any remaining spilled bait. We collected and weighed all spilled bait we could locate and turned over the soil surrounding the bait station to bury any small particles of spilled bait we could not collect.Figure 3Example of activated deterrent devices (scare dancers) mounted above bait stations containing a sodium nitrite toxic bait in north-central Texas, USA during July 2020. Photo property of USDA.Full size imageWe conducted systematic carcass searches along transects following the SN-toxic bait deployment. Specifically, we searched 400 m × 400 m transect grids centered on the bait sites every 50 m, walking transects oriented North/South the first day and East/West the second day. We generated the transects in ArcGIS (v10.8.1, Environmental Systems Research Institute, Redlands, CA, USA), and uploaded them to handheld devices (i.e., mobile phones or tablets) using ArcGIS Explorer (v20.0.1) to navigate along the transects. Additionally, we searched a smaller 50 m × 50 m transect grid centered on the bait sites every 5 m for three consecutive days, again switching between North/South, East/West, and North/South orientation each day, respectively. Transect spacing and distances were based on locations of carcasses found in a previous study with SN-toxic bait24. We searched transects for multiple days to ensure any carcasses were located and to determine if any animals succumbed to consuming spilled SN-toxic bait that may have been missed during our clean-up process days after deployment.We recorded sex, age based on tooth eruption44, weight, location, and evidence of SN-toxic bait consumption of any dead wild pigs that we located. Bait consumption was determined by observing bait in the mouth or stomach, or based on the percentage of methemoglobin in the blood by comparing the red-color-value of a drop of blood on a white laminated card to a standard curve45. For any non-target animals found dead, we recorded species, location, and evidence of SN-toxic bait consumption (as described above).We processed all time-lapse imagery from each bait station using the Colorado Parks and Wildlife Photo Database46. For each image, a single observer recorded the count of each species present. We did not include cattle because they were excluded from bait sites. We used two indices from the images for comparing the rates of visitation by different species. First, we used an index of the count of non-target animals/image during the hours that the deterrent devices were operating (0520–1200 h). We compared this index among the days of pre-, during, and post-activation periods of the deterrent devices to assess if the devices influenced the rate of visitation using linear models in program R. We analyzed sites with and without the deterrent devices separately to assess the effects of each treatment throughout the days independently.For the second index, we estimated rates of the number of wild pigs, non-target mammals, and non-target birds, respectively, observed per hour that visited bait sites. We followed methodology established by22, and used negative binomial generalized mixed models with package glmmTMB37 to compare rates of visitation between periods of pre- and post-SN-toxic bait deployment to assess changes relative to toxic baiting. We considered the change in rates of visitation to be attributed to lethality from SN-toxic bait for the populations of animals visiting the bait sites. We expect this methodology met the assumption that detection of animals remained consistent47 at bait sites because pre- and post-toxic periods were only separated by a single 24-h period when the toxic bait was deployed, and we refreshed the bait daily. We also compared the indices between treatments (with vs without deterrents) and the interaction of period × treatment. The models examined for each group of species were: rate of hourly visitation ~ period + treatment + period × treatment. We also used Site ID as random effects to account for repeated measures taken at each bait site.For the transect analysis, we calculated descriptive summaries of sexes, ages, and distances from carcass to nearest bait station for wild pigs that succumbed to the SN-toxic bait. We also summarized any non-target deaths and distances from the nearest bait site. All research methods for all phases of this study were approved under the USDA National Wildlife Research Center, Institutional Animal Care and Use Committee (protocol QA-3068), and performed and reported in accordance with ARRIVE guidelines and US EPA regulations. More

1.Hein, J. R., Mizell, K., Koschinsky, A. & Conrad, T. A. Deep-ocean mineral deposits as a source of critical metals for high- and green-technology applications: Comparison with land-based resources. Ore Geol. Rev. 51, 1–14 (2013).Article 

Google Scholar 
2.Petersen, S. et al. News from the seabed—Geological characteristics and resource potential of deep-sea mineral resources. Mar. Policy 70, 175–187 (2016).Article 

Google Scholar 
3.Dutkiewicz, A., Judge, A. & Müller, R. D. Environmental predictors of deep-sea polymetallic nodule occurrence in the global ocean. Geology 48, 293–297 (2020).ADS 
CAS 
Article 

Google Scholar 
4.Verlaan, P. A. & Cronan, D. S. Origin and variability of resource-grade marine ferromanganese nodules and crusts in the Pacific Ocean: A review of biogeochemical and physical controls. Geochemistry https://doi.org/10.1016/j.chemer.2021.125741 (2021).Article 

Google Scholar 
5.Radziejewska, T. & Stoyanova, V. Abyssal epibenthic megafauna of the Clarion-Clipperton area (NE Pacific): Changes in time and space versus anthropogenic environmental disturbance. Oceanol. Stud. 29, 83–101 (2000).
Google Scholar 
6.Vanreusel, A., Hilario, A., Ribeiro, P. A., Menot, L. & Arbizu, P. M. Threatened by mining, polymetallic nodules are required to preserve abyssal epifauna. Sci. Rep. 6, 1–6 (2016).Article 
CAS 

Google Scholar 
7.Simon-Lledó, E. et al. Ecology of a polymetallic nodule occurrence gradient: Implications for deep-sea mining. Limnol. Oceanogr. 64, 1883–1894 (2019).ADS 
PubMed 
PubMed Central 
Article 

Google Scholar 
8.Washburn, T. W. et al. Patterns of macrofaunal biodiversity across the Clarion-Clipperton zone: An area targeted for seabed mining. Front. Mar. Sci. 8, 626571 (2021).Article 

Google Scholar 
9.Bonifácio, P., Martinez Arbizu, P. & Menot, L. Alpha and beta diversity patterns of polychaete assemblages across the nodule province of the eastern Clarion-Clipperton Fracture Zone (equatorial Pacific). Biogeosciences 17, 865–886 (2020).ADS 
Article 

Google Scholar 
10.Ansari, Z. A. Distribution of deep-sea benthos in the proposed mining area of Central Indian Basin. Mar. Georesour. Geotechnol. 18, 201–207 (2000).Article 

Google Scholar 
11.Pasotti, F. et al. A local scale analysis of manganese nodules influence on the Clarion-Clipperton Fracture Zone macrobenthos. Deep Sea Res. Part Oceanogr. Res. Pap. 168 (2021).12.Hauquier, F. et al. Geographic distribution of free-living marine nematodes in the Clarion-Clipperton Zone: Implications for future deep-sea mining scenarios. Biogeosciences 16, 3475–3489 (2019).ADS 
CAS 
Article 

Google Scholar 
13.Kuhn, T., Uhlenkott, K., Vink, A., Rühlemann, C. & Martinez Arbizu, P. Manganese nodule fields from the Northeast Pacific as benthic habitats. In Seafloor Geomorphology as Benthic Habitat 2nd edn (eds Harris, P. T. & Baker, E.) 933–947 (Elsevier, 2020). https://doi.org/10.1016/B978-0-12-814960-7.00058-0.Chapter 

Google Scholar 
14.Miljutina, M. A., Miljutin, D. M., Mahatma, R. & Galéron, J. Deep-sea nematode assemblages of the Clarion-Clipperton Nodule Province (Tropical North-Eastern Pacific). Mar. Biodivers. 40, 1–15 (2010).Article 

Google Scholar 
15.Mahatma, R. Meiofauna Communities of the Pacific Nodule Province: Abundance, Diversity and Community Structure (University of Oldenburg, 2009).
Google Scholar 
16.Singh, R. et al. Nematode communities inhabiting the soft deep-sea sediment in polymetallic nodule fields: Do they differ from those in the nodule-free abyssal areas?. Mar. Biol. Res. 12, 1–15 (2016).Article 

Google Scholar 
17.Thiel, H., Schriever, G., Bussau, C. & Borowski, C. Manganese nodule crevice fauna. Deep Sea Res. Part Oceanogr. Res. Pap. 40, 419–423 (1993).ADS 
Article 

Google Scholar 
18.Bussau, C., Schriever, G. & Thiel, H. Evaluation of abyssal metazoan meiofauna from a manganese nodule area of the Eastern South Pacific. Vie Milieu 45, 39–48 (1995).
Google Scholar 
19.Oebius, H. U., Becker, H. J., Rolinski, S. & Jankowski, J. A. Parametrization and evaluation of marine environmental impacts produced by deep-sea manganese nodule mining. Deep Sea Res. Part II Top. Stud. Oceanogr. 48, 3453–3467 (2001).ADS 
CAS 
Article 

Google Scholar 
20.Levin, L. A. et al. Defining “serious harm” to the marine environment in the context of deep-seabed mining. Mar. Policy 74, 245–259 (2016).Article 

Google Scholar 
21.Global Sea Mineral Resources. Environmental Impact Statement—Small-scale testing of nodule collector components on the seafloor of the Clarion-Clipperton Fracture Zone and its environmental impact. 337 (2018).22.Durden, J. M. et al. A procedural framework for robust environmental management of deep-sea mining projects using a conceptual model. Mar. Policy 84, 193–201 (2017).Article 

Google Scholar 
23.Jones, D. O. B. et al. Biological responses to disturbance from simulated deep-sea polymetallic nodule mining. PLoS One 12, e0171750 (2017).ADS 
PubMed 
PubMed Central 
Article 
CAS 

Google Scholar 
24.Jones, D. O. B., Ardron, J. A., Colaço, A. & Durden, J. M. Environmental considerations for impact and preservation reference zones for deep-sea polymetallic nodule mining. Mar. Policy https://doi.org/10.1016/j.marpol.2018.10.025 (2018).Article 

Google Scholar 
25.Boschen, R. E. et al. A primer for use of genetic tools in selecting and testing the suitability of set-aside sites protected from deep-sea seafloor massive sulfide mining activities. Ocean Coast. Manag. 122, 37–48 (2016).Article 

Google Scholar 
26.Boucher, G. & Lambshead, P. J. D. Ecological biodiversity of marine nematodes in samples from temperate, tropical and deep-sea regions. Conserv. Biol. 9, 1594–1604 (1995).Article 

Google Scholar 
27.Ramirez-Llodra, E. et al. Deep, diverse and definitely different: Unique attributes of the world’s largest ecosystem. Biogeosciences 7, 2851–2899 (2010).ADS 
Article 

Google Scholar 
28.Rex, M. A. & Etter, R. J. Deep-Sea Biodiversity: Pattern and Scale (Harvard University Press, 2010).
Google Scholar 
29.Paterson, G. L. J. et al. Biogeography and connectivity in deep-sea habitats with mineral resource potential: A gap analysis. Deliverable 4.2. MIDAS (2014).30.Christodoulou, M., O’Hara, T. D., Hugall, A. F. & Arbizu, P. M. Dark ophiuroid biodiversity in a prospective abyssal mine field. Curr. Biol. 29, 3909–3912 (2019).PubMed 
CAS 
Article 

Google Scholar 
31.Amon, D. J. et al. Insights into the abundance and diversity of abyssal megafauna in a polymetallic-nodule region in the eastern Clarion-Clipperton Zone. Sci. Rep. 6, 30492 (2016).ADS 
PubMed 
PubMed Central 
CAS 
Article 

Google Scholar 
32.Goineau, A. & Gooday, A. J. Diversity and spatial patterns of foraminiferal assemblages in the eastern Clarion-Clipperton zone (abyssal eastern equatorial Pacific). Deep Sea Res. Part Oceanogr. Res. Pap. 149, 103036 (2019).Article 

Google Scholar 
33.Macheriotou, L., Rigaux, A., Derycke, S. & Vanreusel, A. Phylogenetic clustering and rarity imply risk of local species extinction in prospective deep-sea mining areas of the Clarion-Clipperton Fracture Zone. Proc. R. Soc. B Biol. Sci. 287, 20192666 (2020).Article 

Google Scholar 
34.Błażewicz, M., Jóźwiak, P., Menot, L. & Pabis, K. High species richness and unique composition of the tanaidacean communities associated with five areas in the Pacific polymetallic nodule fields. Prog. Oceanogr. 176, 102141 (2019).Article 

Google Scholar 
35.Janssen, A. et al. A reverse taxonomic approach to assess macrofaunal distribution patterns in abyssal pacific polymetallic nodule fields. PLoS One 10, e0117790 (2015).PubMed 
PubMed Central 
Article 
CAS 

Google Scholar 
36.Soetaert, K. & Heip, C. Sample-size dependence of diversity indexes and the determination of sufficient sample size in a high-diversity deep-sea environment. Mar. Ecol. Prog. Ser. 59, 305–307 (1990).ADS 
Article 

Google Scholar 
37.Rose, A. et al. A method for comparing within-core alpha diversity values from repeated multicorer samplings, shown for abyssal Harpacticoida (Crustacea: Copepoda) from the Angola Basin. Org. Divers. Evol. 5, 3–17 (2005).Article 

Google Scholar 
38.George, K. H. et al. Community structure and species diversity of Harpacticoida (Crustacea: Copepoda) at two sites in the deep sea of the Angola Basin (Southeast Atlantic). Org. Divers. Evol. 14, 57–73 (2014).Article 

Google Scholar 
39.Mouillot, D. et al. Rare species support vulnerable functions in high-diversity ecosystems. PLoS Biol. 11, e1001569 (2013).PubMed 
PubMed Central 
CAS 
Article 

Google Scholar 
40.Naeem, S. Species redundancy and ecosystem reliability. Conserv. Biol. 12, 39–45 (1998).Article 

Google Scholar 
41.Turner, P. J., Campbell, L. M. & Van Dover, C. L. Stakeholder perspectives on the importance of rare-species research for deep-sea environmental management. Deep Sea Res. Part Oceanogr. Res. Pap. 125, 129–134 (2017).ADS 
Article 

Google Scholar 
42.Drury, W. H. Rare species. Biol. Conserv. 6, 162–169 (1974).Article 

Google Scholar 
43.Anderson, M. J., Gorley, R. N. & Clarke, K. R. PERMANOVA+ for PRIMER: Guide for Software and Statistical Methods (Primer-E Ltd, 2008).
Google Scholar 
44.Gollner, S. et al. Resilience of benthic deep-sea fauna to mining activities. Mar. Environ. Res. https://doi.org/10.1016/j.marenvres.2017.04.010 (2017).Article 
PubMed 

Google Scholar 
45.Glover, A. G. et al. Polychaete species diversity in the central Pacific abyss: Local and regional patterns, and relationships with productivity. Mar. Ecol. Prog. Ser. 240, 157–170 (2002).ADS 
Article 

Google Scholar 
46.Rosli, N., Leduc, D., Rowden, A. & Robert, K. Review of recent trends in ecological studies of deep-sea meiofauna, with focus on patterns and processes at small to regional spatial scales. Mar. Biodivers. 18, 13–34 (2018).Article 

Google Scholar 
47.Gallucci, F., Moens, T. & Fonseca, G. Small-scale spatial patterns of meiobenthos in the Arctic deep sea. Mar. Biodivers. 39, 9–25 (2009).Article 

Google Scholar 
48.Wieser, W. Die Beziehung zwischen Mundhöhlengestalt, Ernährungsweise und Vorkommen bei freilebenden marinen Nematoden Eine ökologisch-morphologische Studie. Ark. För Zool. 4, 439–483 (1953).
Google Scholar 
49.Leduc, D. Description of Oncholaimus moanae sp. nov. (Nematoda: Oncholaimidae), with notes on feeding ecology based on isotopic and fatty acid composition. J. Mar. Biol. Assoc. U. K. 89, 337–344 (2008).Article 
CAS 

Google Scholar 
50.Pape, E., van Oevelen, D., Moodley, L., Soetaert, K. & Vanreusel, A. Nematode feeding strategies and the fate of dissolved organic matter carbon in different deep-sea sedimentary environments. Deep Sea Res. Part Oceanogr. Res. Pap. 80, 94–110 (2013).ADS 
CAS 
Article 

Google Scholar 
51.Schuelke, T., Pereira, T. J., Hardy, S. M. & Bik, H. M. Nematode-associated microbial taxa do not correlate with host phylogeny, geographic region or feeding morphology in marine sediment habitats. Mol. Ecol. 27, 1930–1951 (2018).PubMed 
Article 

Google Scholar 
52.Tully, B. J. & Heidelberg, J. F. Microbial communities associated with ferromanganese nodules and the surrounding sediments. Extreme Microbiol. 4, 161 (2013).
Google Scholar 
53.Blöthe, M. et al. Manganese-cycling microbial communities inside deep-sea manganese nodules. Environ. Sci. Technol. 49, 7692–7700 (2015).ADS 
PubMed 
Article 
CAS 

Google Scholar 
54.Maybury, C. Crevice Foraminifera from abyssal South East Pacific manganese nodules. In Microfossils and Oceanic Environments (eds Moguilevsky, A. & Whatley, R.) (University of Wales, 1996).
Google Scholar 
55.Pape, E., Bezerra, T. N., Hauquier, F. & Vanreusel, A. Limited spatial and temporal variability in meiofauna and nematode communities at distant but environmentally similar sites in an area of interest for deep-sea mining. Front. Mar. Sci. 4, 205 (2017).Article 

Google Scholar 
56.Uhlenkott, K., Vink, A., Kuhn, T. & Arbizu, P. M. Meiofauna in a potential deep-sea mining area—Influence of temporal and spatial variability on small-scale abundance models. Diversity 13, 3 (2021).CAS 
Article 

Google Scholar 
57.Veillette, J., Juniper, S. K., Gooday, A. J. & Sarrazin, J. Influence of surface texture and microhabitat heterogeneity in structuring nodule faunal communities. Deep Sea Res. Part Oceanogr. Res. Pap. 54, 1936–1943 (2007).ADS 
Article 

Google Scholar 
58.Tilot, V., Ormond, R., Moreno Navas, J. & Catalá, T. S. The Benthic Megafaunal Assemblages of the CCZ (Eastern Pacific) and an approach to their management in the face of threatened anthropogenic impacts. Front. Mar. Sci. 5, 7 (2018).Article 

Google Scholar 
59.ISA. Recommendations for the guidance of contractors for the assessment of the possible environmental impacts arising from exploration for marine minerals in the Area (2020).60.ISA. Draft regulations on exploitation of mineral resources in the Area (2019).61.ISA. Environmental Management Plan for the Clarion-Clipperton Zone (2011).62.Wedding, L. M. et al. From principles to practice: A spatial approach to systematic conservation planning in the deep sea. Proc. R. Soc. B Biol. Sci. 280, 20131684 (2013).CAS 
Article 

Google Scholar 
63.ISA. Deep CCZ Biodiversity Synthesis Workshop Report. 206 (2020).64.McQuaid, K. A. et al. Using habitat classification to assess representativity of a protected area network in a large, data-poor area targeted for deep-sea mining. Front. Mar. Sci. 7, 558860 (2020).Article 

Google Scholar 
65.Mullineaux, L. S. The role of settlement in structuring a hard-substratum community in the deep sea. J. Exp. Mar. Biol. Ecol. 120, 247–261 (1988).Article 

Google Scholar 
66.Cuvelier, D. et al. Potential mitigation and restoration actions in ecosystems impacted by seabed mining. Front. Mar. Sci. 5, 467 (2018).Article 

Google Scholar 
67.De Smet, B. et al. The community structure of deep-sea macrofauna associated with polymetallic nodules in the eastern part of the Clarion-Clipperton fracture zone. Front. Mar. Sci. 4, 103 (2017).
Google Scholar 
68.Bezerra, T. N. et al. Nemys: World Database of Nematodes. http://nemys.ugent.be. https://doi.org/10.14284/366 (2021).69.George, K.-H. Gemeinschaftsanalytische Untersuchungen der Harpacticoidenfauna der Magellanregion, sowie erste similaritätsanalytische Vergleiche mit Assoziationen aus der Antarktis = Community analysis of the harpacticoid fauna of the Magellan Region, as well as first comparisons with antarctic associations, based on similarity analyses. Berichte Zur Polarforsch. Rep. Polar Res. 327, 1–187 (1999).
Google Scholar 
70.Moens, T. & Vincx, M. Observations on the feeding ecology of estuarine nematodes. J. Mar. Biol. Assoc. U. K. 77, 211–227 (1997).Article 

Google Scholar 
71.Guilini, K., Van Oevelen, D., Soetaert, K., Middelburg, J. J. & Vanreusel, A. Nutritional importance of benthic bacteria for deep-sea nematodes from the Arctic ice margin: Results of an isotope tracer experiment. Limnol. Oceanogr. 55, 1977–1989 (2010).ADS 
CAS 
Article 

Google Scholar 
72.Clarke, K. & Gorley, R. PRIMER v6: User Manual/Tutorial (Primer-E Ltd, 2006).
Google Scholar 
73.R Core Team. R: A Language and Environment for Statistical Computing (R Foundation for Statistical Computing, 2019).
Google Scholar 
74.Wickham, H. ggplot2: Elegant Graphics for Data Analysis (Springer, 2016).MATH 
Book 

Google Scholar 
75.Wilke, C. O. cowplot: Streamlined Plot Theme and Plot Annotations for ‘ggplot2’ (2019).76.Oksanen, J. et al. vegan: Community Ecology Package (2019).77.Martinez Arbizu, P. M. pairwiseAdonis: Pairwise Multilevel Comparison using Adonis (2017).78.Chao, A. et al. Rarefaction and extrapolation with Hill numbers: A framework for sampling and estimation in species diversity studies. Ecol. Monogr. 84, 45–67 (2014).Article 

Google Scholar 
79.Hsieh, T. C. & Chao, A. Package iNEXT 2.0.19: Interpolation and extrapolation of species diversity (2019).80.Schenker, N. & Gentleman, J. F. On judging the significance of differences by examining the overlap between confidence intervals. Am. Stat. 55, 182–186 (2001).MathSciNet 
Article 

Google Scholar 
81.Gehlenborg, N. UpSetR: A More Scalable Alternative to Venn and Euler Diagrams for Visualizing Intersecting Sets (2019).82.Simpson, G. L. permute: Functions for Generating Restricted Permutations of Data (2019).83.Baselga, A., Orme, D., Villeger, S., Bortoli, J. D. & Leprieur, F. betapart: Partitioning Beta Diversity into Turnover and Nestedness Components (2018). More

Sequencing of Shorea leprosula genomeSample collectionLeaf samples of S. leprosula were obtained from a reproductively mature (diameter at breast height, 50 cm) diploid tree B1_19 (DNA ID 214) grown in the Dipterocarp Arboretum, Forest Research Institute Malaysia (FRIM).DNA extractionGenomic DNA was extracted from leaf samples using the 2% cetyltrimethylammonium bromide (CTAB) method90 and purified using a High Pure PCR Template Purification kit (Roche).Library preparation and sequencingPaired-end (170, 500, and 800 bp) and mate-pair (2 kb) genomic libraries were prepared using a TruSeq DNA Library Preparation kit (Illumina) and a Mate Pair Library Preparation kit (Illumina), respectively. Mate-pair libraries with larger insert sizes were constructed using a Nextera Mate Pair Library Preparation kit (Illumina). Ten micrograms of genomic DNA were tagmented in a 400 μl reaction and fractionated using SageELF, with the recovery of 11 fractions with 3–16+ kb. Each fraction was circularized and fragmented with a Covaris S2. Biotin-containing fragments were purified using Dynabeads M-280 streptavidin beads. Sequencing adapters (KAPA TruSeq Adapter kit) were attached using a KAPA Hyper Prep kit. The libraries were amplified for 10–13 cycles and purified with 0.8× AMpure XP. DNA libraries were then sequenced (~388× coverage) using Illumina HiSeq2000 (TruSeq libraries) and HiSeq2500 (Nextera libraries) at the Functional Genomics Center Zurich (FGCZ), University of Zurich, Switzerland (Supplementary Table 1).Genome assemblyAdapters and low-quality bases for all paired-end and mate-pair reads were removed using Trimmomatic91. The filtered paired-end reads of the 170 bp library were used to identify the genome size using k-mer distribution generated by Jellyfish92 that was implemented in the scripts by Joseph Ryan42. The raw R1 reads from paired-end 170 and 800 bp libraries (clipped at 95 bp, representing about 70 genome equivalents) were used to estimate the heterozygosity using KAT43 with a k-mer size of 23 nt. De novo genome assembly of all reads was performed using ALLPATHSLG assembler v5248840.Assembly verification and assessment of the assembled genomeAssembly validationTo validate the genome assembly, we mapped (i) the short reads used for the genome assembly, (ii) scanned the assembly for the presence of single-copy orthologs, and (iii) mapped transcriptome sequences obtained from seven organs.Assembly verification by mapping of short readsFor each library used for genome assembly, all trimmed reads were aligned to the assembled S. leprosula genome using Burrows–Wheeler Aligner (BWA) v0.7.1293. Then, mapping ratio was calculated for each BAM file using Samtools94 with “flagstat” command.Identification of highly conserved single-copy orthologsBUSCO v3.1.042 was run with the Embryophyta dataset and Arabidopsis as the species for AUGUSTUS prediction (see subsection below “Protein-coding gene prediction”).Assembly verification by mapping transcriptome sequencesFor mapping transcriptome sequences, samples of seven organs (leaf bud, flower bud, flower, inner bark, small seed, large seed, and calyx) were obtained from the S. leprosula individual used for the genome sequencing (Supplementary Table 2). Total RNA was extracted from each sample using RNeasy Plant Mini Kit (Qiagen) and it was treated with Turbo DNase I (Takara). Library preparation was carried out using a TruSeq RNA Library Preparation kit (Illumina). Paired-end sequencing was conducted for all the libraries using Illumina HiSeq2000 at the FGCZ, University of Zurich, Switzerland. Adapters and low-quality bases for all paired-end reads were removed using Trimmomatic. The trimmed sequences of each library were mapped onto the assembled genome using STAR aligner v2.4.2a95, and mapping ratio was obtained from the output file of STAR.Genome annotationRepeat sequence analysisBoth homology-based and de novo prediction analyses were used to identify the repeat content in the S. leprosula assembly. For the homology-based analysis, we used Repbase (version 20120418) to perform a TE search with RepeatMasker (4.0.5) and the WuBlast search engine. For the de novo prediction analysis, we used RepeatModeler to construct a TE library. Elements within the library were then classified by homology to Repbase sequences (see subsection below “Preparation of repeat sequences for evidence-based gene prediction”).Protein-coding gene predictionS. leprosula protein-coding genes were predicted by AUGUSTUS v3.245. For ab initio gene prediction, we used a pre-trained A. thaliana metaparameter implemented in AUGUSTUS. For the evidence-based gene prediction, we used the information of exon, intron and repeat sequences of S. leprosula as hints for the AUGUSTUS gene prediction. The details of the preparation of the hints were described in the following subsections.Preparation of repeat sequences for evidence-based gene predictionWe used RepeatModeler to construct a de novo library of repeated sequences in the S. leprosula assembly. Then, using RepeatMasker, we generated a file containing the information of the positions of repeat sequences in the S. leprosula genome based on the RepeatModeler library. Elements within the library were then classified by homology to Repbase sequences. Finally, the hint file for repeat sequences in GFF format was prepared using the two scripts, “10_makeGffRm.pl” and “12_makeTeHints.pl”, stored in https://gitlab.com/rbrisk/ahalassembly.Preparation of the exon and intron information for evidence-based gene predictionTo obtain the exon and intron hints, we used the mapping data of RNA-seq obtained from seven organs of the sequenced S. leprosula individual as described above. First, we merged all the mapping data stored in different BAM files into a single BAM file using SAMtools. Then, we prepared the intron hint file in GFF format using the, “bam2hints” script of AUGUSTUS. The exon hint file was also generated from the merged BAM file using the two AUGUSTUS scripts, “bam2wig” and “wig2hints.pl”. To conduct evidence-based gene prediction with AUGUSTUS, the three hint files (repeat sequences, intron and exon) described above were merged into a single file in GFF format.BUSCO analysisGenome annotation completeness were assessed with BUSCO v3.1.044 using the Embryophyta odb9 dataset composed of 1440 universal Embryophyta single-copy genes. We referred to these 1440 genes as core genes in the main text.Comparison with the proteome of Theobroma cacao
T. cacao’s gene models18 were downloaded from Phytozome 11 (https://phytozome.jgi.doe.gov/pz/portal.html). Then, comparison was conducted with BLASTP96 using the T. cacao proteomes as the BLAST database (E-value cutoff: 1.0E-10). Only the best hit was stored for each gene. We considered these best hits of the T. cacao genes as orthologs of the S. leprosula genes. When the T. cacao orthologs were identified by the BLASTP search, the orthologs of A. thaliana were defined based on the T. cacao-A. thaliana orthologous information provided by Phytozome 11 (Supplementary Table 4). When the T. cacao orthologs were not identified, the orthologs of A. thaliana were searched by BLASTP (E-value cutoff: 1.0E-10) using the A. thaliana proteomes obtained from TAIR 10 (https://www.arabidopsis.org) as the BLAST database.Synteny analysisBased on the result of the above BLASTP searches, we assessed synteny between the S. leprosula scaffolds and the T. cacao chromosomes using MCScanX97. Genome information of T. cacao in GFF format was also obtained from Phytozome 11 as described above, which was used as an input file for MCScanX.Assessment of the genome assemblyPopulation data and other dipterocarp speciesTo assess whether the genome assembly could be used as a reference for the S. leprosula individuals from various populations, we checked mapping ratio, SNP positions, and admixture using the distribution-wide S. leprosula samples. Similarly, to assess whether the S. leprosula assembly could be used as a reference for aligning data from closely related species and determining their mapping ratios. For interspecific analysis, the following three Dipterocarpoideae species: S. platycarpa, D. aromatica, and N. heimii were used (Supplementary Table 7).Sample collection and DNA extractionLeaf samples of 19 S. leprosula individuals from different populations and three other dipterocarp species (S. platycarpa, D. aromatica, and N. heimii) were used as described in Supplementary Tables 6 and 7. Genomic DNA was extracted using the same method as described above.Library preparation and sequencingPaired-end genomic libraries (200 bp) were prepared using a TruSeq DNA Library Preparation kit (Illumina). DNA libraries were then sequenced (~16× coverage each) using Illumina HiSeq2000.Mapping and SNP callingAdapters and low-quality bases from resequencing reads were removed using Trimmomatic. All trimmed reads were then mapped and aligned to the S. leprosula assembly using BWA. Variants were called using GATK v3.598. Duplicated reads were marked using Picard 2.6.0. Within GATK, HaplotypeCaller was used to identify variants for each sample by generating an intermediate genomic variant call format (gVCF). Subsequently, gVCF files were merged using GenotypeGVCFs to produce a raw VCF file containing SNPs and INDELs. Low-quality variants were removed from the raw VCF file by applying the hard filters implemented in GATK. Variants with genotype quality (GQ)  More

Host and gut content samplingA total of 94 individuals (42 A. speciosus, 9 A. argenteus, and 43 M. rufocanus) were captured from four sites within the Kamikawa Chubu national forest in the central area on the island of Hokkaido, Japan (Supplementary Table S1), and a total of 280 gut content (from the small intestine, cecum, and colon) and fecal matter (from the rectum) samples were collected for microbiome analysis (Supplementary Table S2). Based on 16S rRNA amplicon sequencing using Illumina Miseq, a total of 12,286,171 paired-end reads were obtained after quality filtering and chimeric sequence removal. There was an average of 43,879 reads per sample, although it varied among species and gut region (Supplementary Table S3).Within host species/among gut region gut microbiota alpha diversityAlpha diversity of the gut microbiota in the small intestine was significantly lower than the rectum, colon, and cecum in all three host species based on Shannon diversity, Faith’s PD, evenness, and number of ASVs as expected (GLME: all p  0.05; Fig. 1, Supplementary Fig. S2, Supplementary Tables S4–S7). Males had significantly higher alpha diversity within all gut regions of A. speciosus while female A. argenteus had significantly higher alpha diversity as compared to males (GLME, all p  0.05; Supplementary Tables S4–S7) while age had no effect in any gut region of any rodent species (GLME: all p  > 0.05; Supplementary Tables S4–S7).Figure 1Alpha diversity within each gut region of each species based on (a) Shannon diversity and (b) Faith’s PD. Dashed lines separate host species.Full size imageAmong host species alpha diversityMyodes rufocanus had significantly higher alpha diversity in all four gut regions as compared to both A. speciosus and A. argenteus based on all four diversity measurements (GLME: all p  More

Isolation and characterization of bean dreg-degrading strainsA 1362-bp amplification fragment of 16S rDNA was obtained by PCR (GenBank accession number MW406939). This sequence was compared with others in the GenBank database, aligning the 16S rDNA sequences with several Geobacillus sp. strains and constructed a phylogenetic tree (Fig. 2a). The phylogenetic tree clearly showed that strain GT-02 belongs to the G.toebii branch and was similar to G.toebii R-32652, G.toebii NBRC 107807, and G.toebii SK-1 with 99.78%, 99.63% and 99.05% similarities, respectively. According to the study described previously, G.toebii was a gram-positive, aerobic rod and motile bacterial26. G.toebii could produce acid from inositol and gas from nitrate. G.toebii could hydrolysis casein and utilize n-alkanes as carbon source27.Figure 2(a) Phylogenetic tree based on 16S rDNA gene sequences from related species of the genus Geobacillus constructed using the neighbour-joining method with 1000 bootstrap replicates. Branch length is indicated at each node. (b) The growth curve of strain GT-02 with temperature. (c) The growth curve of strain GT-02 with pH.Full size imageThe growth characteristics of strain GT-02, such as temperature and pH values, were investigated. The bacterial strain could grow within a range of 40–75 °C and pH 6.50–9.50, and the optimum temperature and pH were 65 °C and 7.50, respectively (Fig. 2b,c). Compared to other G.toebii strains, the maximum growth temperature and pH of strains R-32652 and SK-1 were 70 °C and 9.0026,28, respectively. These results showed that strain GT-02 was more resistant to high temperature and alkalinity. Fermentation temperature above 70 °C could effectively inactivate harmful microorganisms in organic solid waste12. Therefore, the fermentation temperature was set at 70 °C in this study.Changes in the composition of bean dregs during fermentationChanges in GI, TOC and TN of bean dregs during fermentationThe GI is traditionally used to evaluate the phytotoxicity and maturity of organic fertilizer12. As shown in Fig. 3a, both groups of experiments reached the standard of maturity (GI ≥ 85.00%). Therefore, the fermentation was terminated in five days. In the initial stage of fermentation, the GI of CK dropped to 51.85% on day 2, and the GI of T1 dropped to 41.98% on day 1. Phytotoxicity, which is usually caused by various heavy metals and low-molecular-weight substances, such as NH3 and organic acids, can reduce seed germination and inhibit root development29. During fermentation, bean dregs might produce NH3, organic acids and other substances, which could trigger a decrease in the GI. The GI of T1 showed a clear decrease, which was likely due to the production of toxic organic acids and might also explain the decrease in pH observed in T1 (Fig. 3d). Due to the degradation of organic acids, the GI of T1 increased to 95.06% on the third day and continued to increase to more than 100.00%, whereas in CK, the GI only reached 86.42% at the end of the fermentation. These results revealed that the maturity of T1 on day 3 was markedly higher than that of CK on day 5 and thus suggest that G.toebii can significantly enhance the fermentation efficiency by accelerating the maturation process and thus reducing the thermophilic fermentation period from 5 to 3 days.Figure 3Profiles of GI (a), TOC (b), TN (c), pH (d) and EC (e) during the fermentation process of CK and T1. The data represent the means ± standard deviations from three measurements.Full size imageTOC is usually used as an energy source by microorganisms30. The TOC loss in both CK and T1 increased during fermentation (Fig. 3b). The reduction of TOC was mainly caused by the production of carbon dioxide from bacterial respiration. The rate of TOC loss in T1 was higher than that in CK. At the end of the fermentation, the TOC loss of T1 was 11.78% higher than that in CK. Because of the addition of G.toebii, bacterial metabolism in T1 was more active, and organic degradation was faster.The TN loss in both CK and T1 also showed an upward trend (Fig. 3c). The loss of TN was mainly caused by the volatilization of ammonia nitrogen31. The rate of TN loss in T1 increased more than that of CK group. After fermentation (day 5), the TN loss in T1 was 6.83% higher than that of CK. The mineralization in T1 was more active and thus ammonia nitrogen was more, which was easy to cause volatilization. However, the bean dregs in CK were mature on the 5th day, while those in T1 were on the 3rd day. At this time, the TN loss of mature bean dregs in T1 was 5.66% lower than that in CK, which indicated that the bean dregs lost less nitrogen source when they reached the standard of maturity after the addition of G.toebii.Changes in pH and EC of bean dregs during fermentationThe variation in pH observed during fermentation is due to the interaction between inorganic nitrogen and organic acids produced by the decomposition of organic matter32. As shown in Fig. 3d, the pH of CK gradually increased to 8.72 at the end of the fermentation. The ammonification process and the release of free NH3 during organic matter (OM) degradation lead to increases in pH33. The pH of T1 decreased to 5.73 on day 1, which was due to the formation of more organic acids than CK, and then increased to 8.76 on day 2, which was due to acid consumption and ammonia formation. Figure 2c showed that GT-02 could hardly grow when the pH was lower than 6.00, but the heterogeneity of solid fermentation provided a possible living environment for the growth of GT-02. Subsequently, the pH of T1 slowly decreased to 8.10 due to ammonia volatilization or ammonia conversion. These study findings showed that the pH value of the fermentation process was significantly affected by the addition of GT-02. G.toebii can produce abundant high-temperature enzymes, such as amylase, protease, cellulase, xylanase, and mannanase17, which explains why the ammonification process was faster in T1 than in CK and thus the higher pH was found in T1.The EC, which is a measure of the total ion concentration, describes changes in the levels of organic and inorganic ions such as SO42−, Na+, NH4+, K+, Cl−, and NO3− during the fermentation process34. As shown in Fig. 3e, the EC of the two groups increased significantly during fermentation process (P  More

1.Collins, M. et al. Long-term climate change: Projections, commitments and irreversibility. In Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (Cambridge University Press, 2013).2.Kawahata, H. et al. Perspective of the response by marine calcifiers to global warming and ocean acidification –Behavior of corals and foraminifers in the high CO2 world in “hot house”. Prog. Earth Planet Sci. 6, 5 (2019).Article 

Google Scholar 
3.Kroeker, K. J., Kordas, R. L., Crim, R. N. & Singh, G. G. Meta-analysis reveals negative yet variable effects of ocean acidification on marine organisms. Ecol. Lett. 13, 1419–1434 (2010).Article 

Google Scholar 
4.Orr, J. C. et al. Anthropogenic ocean acidification over the twenty-first century and its impact on calcifying organisms. Nature 437, 681–686 (2005).ADS 
CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
5.Schiebel, R. Planktic foraminiferal sedimentation and the marine calcite budget. Glob. Biogeochem. Cycles 16, 1065 (2002).ADS 
Article 
CAS 

Google Scholar 
6.Keul, N., Langer, G., de Nooijer, L. J. & Bijma, J. Effect of ocean acidification on the benthic foraminifera Ammonia sp. is caused by a decrease in carbonate ion concentration. Biogeosciences 10, 6185–6198 (2013).ADS 
CAS 
Article 

Google Scholar 
7.Doo, S. S., Fujita, K., Byrne, M. & Uthicke, S. Fate of calcifying tropical symbiont-bearing large benthic Foraminifera: Living sands in a changing ocean. Biol. Bull. 226, 169–186 (2014).CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
8.Prazeres, M., Uthicke, S. & Pandolfi, J. M. Ocean acidification induces biochemical and morphological changes in the calcification process of large benthic foraminifera. Proc. R. Soc. B 282, 20142782 (2015).PubMed 
PubMed Central 
Article 
CAS 

Google Scholar 
9.Iwasaki, S. et al. Sensitivity of planktic foraminiferal test bulk density to ocean acidification. Sci. Rep. 9, 9803 (2019).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
10.Hohenegger, J., Kinoshita, S., Briguglio, A., Eder, W. & Wöger, J. Lunar cycles and rainy seasons drive growth and reproduction in nummulitid foraminifera, important producers of carbonate buildups. Sci. Rep. 9, 8286 (2019).ADS 
PubMed 
PubMed Central 
Article 
CAS 

Google Scholar 
11.Kinoshita, S. et al. Temperature effects on the shell growth of a larger benthic foraminifer (Sorites orbiculus): Results from culture experiments and micro X-ray computed tomography. Mar. Micropaleontol. 163, 101960 (2021).ADS 
Article 

Google Scholar 
12.Fujita, K. & Fujimura, H. Organic and inorganic carbon production by algal symbiont-bearing foraminifera on northwest Pacific coral-reef flat. J. Foraminifer. Res. 38, 117–126 (2008).Article 

Google Scholar 
13.Raja, R., Saraswati, P. K., Rogers, K. & Iwao, K. Magnesium and strontium compositions of recent symbiont-bearing benthic foraminifera. Mar. Micropaleontol. 58, 31–44 (2005).ADS 
Article 

Google Scholar 
14.Narayan, G. R. et al. Response of large benthic foraminifera to climate and local changes: Implications for future carbonate production. Sedimentology. 12858. https://doi.org/10.1111/sed.12858 (2021).
15.Morse, J. W., Andersson, A. J. & Mackenzie, F. T. Initial responses of carbonate-rich shelf sediments to rising atmospheric pCO2 and “ocean acidification”: Role of high Mg-calcites. Geochim. Cosmochim. Acta 70, 5814–5830 (2006).ADS 
CAS 
Article 

Google Scholar 
16.Fujita, K., Nishi, H. & Saito, T. Population dynamics of Marginopora kudakajimaensis Gudmundsson (Foraminifera: Soritidae) in the Ryukyu Islands, the tropical northwest Pacific. Mar. Micropaleontol. 38, 267–284 (2000).ADS 
Article 

Google Scholar 
17.Kuroyanagi, A., Kawahata, H., Suzuki, A., Fujita, K. & Irie, T. Impacts of ocean acidification on large benthic foraminifers: Results from laboratory experiments. Mar. Micropaleontol. 73, 190–195 (2009).ADS 
Article 

Google Scholar 
18.Barker, S. & Elderfield, H. Foraminiferal calcification response to glacial–interglacial changes in atmospheric CO2. Science 297, 833–836 (2002).ADS 
CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
19.Osborne, E. B. et al. Calcification of the planktonic foraminifera Globigerina bulloides and carbonate ion concentration: Results from the Santa Barbara Basin. Paleoceanography 31, 1083–1102 (2016).ADS 
Article 

Google Scholar 
20.Mollica, N. R. et al. Ocean acidification affects coral growth by reducing skeletal density. Proc. Natl. Acad. Sci. 115, 1754–1759 (2018).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
21.Schmidt, C., Kucera, M. & Uthicke, S. Combined effects of warming and ocean acidification on coral reef Foraminifera Marginopora vertebralis and Heterostegina depressa. Coral Reefs 33, 805–818 (2014).ADS 
Article 

Google Scholar 
22.Sinutok, S., Hill, R., Kühl, M., Doblin, M. & Ralph, P. Ocean acidification and warming alter photosynthesis and calcification of the symbiont-bearing foraminifera Marginopora vertebralis. Mar. Biol. 161, 2143–2154 (2014).CAS 
Article 

Google Scholar 
23.ter Kuile, B., Erez, J. & Padan, R. Mechanisms for the uptake of inorganic carbon by two species of symbiont-bearing foraminifera. Mar. Biol. 103, 241–251 (1989).Article 

Google Scholar 
24.Nijweide, P. J., Kawilarang-de Haas, E. W. & Wassenaar, A. M. Alkaline phosphatase and calcification, correlated or not?. Metab. Bone Dis. Relat. Res. 3, 61–66 (1981).CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
25.Guo, M. K. & Messer, H. H. A comparison of Ca2+-, Mg2+-ATPase and alkaline phosphatase activities of rat incisor pulp. Calc. Tissue Res. 26, 33–38 (1978).CAS 
Article 

Google Scholar 
26.Vogel, N. & Uthicke, S. Calcification and photobiology in symbiont-bearing benthic foraminifera and responses to a high CO2 environment. J. Exp. Mar. Biol. Ecol. 424–425, 15–24 (2012).Article 
CAS 

Google Scholar 
27.Schiebel, R. & Hemleben, C. Planktic Foraminifers in the Modern Ocean (Springer, 2017).Book 

Google Scholar 
28.Bassinot, F. C., Mélières, F., Gehlen, M., Levi, C. & Labeyrie, L. Crystallinity of foraminifera shells: A proxy to reconstruct past botto m water CO3= changes?. Geochem. Geophys. Geosyst. 5, Q08D10 (2004).Article 

Google Scholar 
29.Broecker, W. & Clark, E. Shell weights from the South Atlantic. Geochem. Geophys. Geosyst. 5, Q03003 (2004).ADS 
Article 

Google Scholar 
30.Beer, C. J., Schiebel, R. & Wilson, P. A. Testing planktic foraminiferal shell weight as a surface water [CO32−] proxy using plankton net samples. Geology 38, 103–106 (2010).ADS 
CAS 
Article 

Google Scholar 
31.Naik, S. S., Naidu, P. D., Govil, P. & Godad, S. Relationship between weights of planktonic foraminifer shell and surface water CO3= concentration during the Holocene and Last Glacial Period. Mar. Geol. 275, 278–282 (2010).ADS 
CAS 
Article 

Google Scholar 
32.Moy, A. D., Howard, W. R., Bray, S. G. & Trull, T. W. Reduced calcification in modern Southern Ocean planktonic foraminifera. Nat. Geosci. 2, 276–280 (2009).ADS 
CAS 
Article 

Google Scholar 
33.Gonzalez-Mora, B., Sierro, F. J. & Flores, J. A. Controls of shell calcification in planktonic foraminifers. Quat. Sci. Rev. 27, 956–961 (2008).ADS 
Article 

Google Scholar 
34.Marr, J. P. et al. Ecological and temperature controls on Mg/Ca ratios of Globigerina bulloides from the southwest Pacific Ocean. Paleoceanography 26, PA2209 (2011).ADS 
Article 

Google Scholar 
35.de Villiers, S. A 425 ka record of foraminiferal shell weight variability in the western Equatorial Pacific. Paleoceanography 18, 1080 (2003).ADS 

Google Scholar 
36.de Villiers, S. Occupation of an ecological niche as the fundamental control on the shell-weight of calcifying planktonic foraminifera. Mar. Biol. 144, 45–50 (2004).Article 

Google Scholar 
37.Reymond, C. E., Lloyd, A., Kline, D. I., Dove, S. G. & Pandolfi, J. M. Decline in growth of foraminifer Marginopora rossi under eutrophication and ocean acidification scenarios. Glob. Change Biol. 19, 291–302 (2013).ADS 
Article 

Google Scholar 
38.Weinkauf, M. F. G., Moller, T., Koch, M. C. & Kucera, M. Calcification intensity in planktic foraminifera reflects ambient conditions irrespective of environmental stress. Biogeosciences 10, 6639–6655 (2013).ADS 
CAS 
Article 

Google Scholar 
39.Doo, S. S. et al. Amelioration of ocean acidification and warming effects through physiological buffering of a macroalgae. Ecol. Evol. 10, 8465–8475 (2020).PubMed 
PubMed Central 
Article 

Google Scholar 
40.Hikami, M. et al. Contrasting calcification responses to ocean acidification between two reef foraminifers harboring different algal symbionts. Geophys. Res. Lett. 38, L19601 (2011).ADS 
Article 
CAS 

Google Scholar 
41.Sanyal, A. et al. Oceanic pH control on the boron isotopic composition of foraminifera: Evidence from culture experiments. Paleoceanography 11, 513–517 (1996).ADS 
Article 

Google Scholar 
42.Anagnostou, E. et al. Changing atmospheric CO2 concentration was the primary driver of early Cenozoic climate. Nature 533, 380–384 (2016).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
43.Foster, G. L. & Rae, J. W. B. Reconstructing ocean pH with boron isotopes in foraminifera. Annu. Rev. Earth Planet. Sci. 44, 207–237 (2016).ADS 
CAS 
Article 

Google Scholar 
44.Kroeker, K. J. et al. Impacts of ocean acidification on marine organisms: Quantifying sensitivities and interaction with warming. Glob. Change Biol. 19, 1884–1896 (2013).ADS 
Article 

Google Scholar 
45.Dove, S. G. et al. Future reef decalcification under a business-as-usual CO2 emission scenario. Proc. Nat. Acad. Sci. 110, 15342–15347 (2013).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
46.Cornwall, C. E. et al. Global declines in coral reef calcium carbonate production under ocean acidification and warming. Proc. Nat. Acad. Sci. 118, 2015265118 (2021).Article 
CAS 

Google Scholar 
47.Langer, M. R., Silk, M. T. & Lipps, J. H. Global ocean carbonate and carbon dioxide production: the role of reef foraminifera. J. Foraminifer. Res 27, 271–277 (1997).Article 

Google Scholar 
48.Pierrot, D., Lewis E. D. & Wallace, D.W. MS EXCEL Program Developed for CO2 System Calculations. ORNL/CDIAC-105a. (Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, U.S. Department of Energy, 2006). https://doi.org/10.3334/cdiac/otg.co2sys_xls_cdiac105a.49.Shapiro, S. S. & Wilk, M. B. An analysis of variance test for normality (complete samples). Biometrika 52, 591–611 (1965).MathSciNet 
MATH 
Article 

Google Scholar 
50.Bartlett, M. S. Properties of sufficiency and statistical test. Proc. R. Soc. A 160, 268–282 (1937).ADS 
MATH 

Google Scholar  More

Infection-age structured dynamicsFor the description of the dynamics, we follow the customary infection-age structured approach (for details see for instance Refs.4,10,11,12). Explicitly, we consider the infection-age structured dynamics of the number of individuals ({u}_{I}left(t,tau right)) at time (t) who were infected at time (t-tau) given by$$begin{array}{c}frac{partial }{partial t}{u}_{I}left(t,tau right)+frac{partial }{partial tau }{u}_{I}left(t,tau right)=0end{array}$$
(7)
with boundary condition$$begin{array}{c}{u}_{I}left(t,0right)=jleft(tright).end{array}$$
(8)
Here, (tau) is the time elapsed after infection, referred to as infection age, and (jleft(tright)={int }_{0}^{infty }{k}_{I}(t,tau ){u}_{I}left(t,tau right)dtau) is the incidence, with ({k}_{I}(t,tau )) being the rate of secondary transmissions per single primary case.The solution is obtained through the method of characteristics32 as$$begin{array}{c}{u}_{I}left(t,tau right)=jleft(t-tau right)end{array}$$
(9)
for (tge tau) and ({u}_{I}left(t,tau right)=0) for (t1 for countries and for US locations.The daily death counts (Delta {n}_{W}left(tright)={n}_{W}left(tright)-{n}_{W}left(t-1right)) are considered to contain reporting artifacts if they are negative or if they are unrealistically large. This last condition is defined explicitly as larger than 4 times its previous 14-day average value plus 10 deaths, (Delta {n}_{W}left(tright) >10+4times frac{1}{14}left({n}_{W}left(tright)-{n}_{W}left(t-14right)right)), from a non-sparse reporting schedule with at least 2 consecutive non-zero values before and after the time (t), (Delta {n}_{W}left(tright)ne frac{1}{5}left({n}_{W}left(t+2right)-{n}_{W}left(t-3right)right)).Reporting artifacts identified at time (t) are considered to be the result of previous miscounting. The excess or lack of deaths are imputed proportionally to previous death counts. Explicitly, death counts are updated as$$begin{array}{c}{n}_{W}left(t-1-iright)leftarrow {n}_{W}left(t-1-iright)frac{{n}_{W}{left(t-1right)}_{estimated}}{{n}_{W}left(t-1right)}end{array}$$
(32)
with ({n}_{W}{left(t-1right)}_{estimated}={n}_{W}left(tright)-frac{1}{7}left({n}_{W}left(t-1right)-{n}_{W}left(t-8right)right)) for all (ige 0). In this way, (Delta {n}_{W}left(tright)) is assigned its previous seven-day average value.The expected daily deaths, (Delta {n}_{D}(t)), are obtained through a density estimation multiscale functional, ({f}_{de}), as (Delta {n}_{D}(t)={f}_{de}left(Delta {n}_{W}left(tright)right)), which leads to the estimation of the expected cumulative deaths at time (t) as ({n}_{D}left(tright)={n}_{W}left({t}_{0}right)+{sum }_{s={t}_{0}+1}^{t}Delta {n}_{D}(s)). Specifically,$$begin{array}{c}{f}_{de}left(Delta {n}_{W}left(tright)right)=left(1-{r}_{1}right)d{d}_{0}+{r}_{1}left(left(1-{r}_{2}right)d{d}_{1}+{r}_{2}d{d}_{2}right)end{array}$$
(33)
with$$begin{array}{c}{r}_{1} = {e}^{-0.3d{d}_{1}},end{array}$$
(34)
$$begin{array}{c}{r}_{2} = {e}^{-3d{d}_{2}},end{array}$$
(35)
$$begin{array}{c}d{d}_{0}={ma}_{14}left({ma}_{14}left(Delta {n}_{W}left(tright)right)right),end{array}$$
(36)
$$begin{array}{c}d{d}_{1}={rg}_{12}left({ma}_{14}left(Delta {n}_{W}left(tright)right)right),end{array}$$
(37)
$$begin{array}{c}d{d}_{2}={rg}_{48}left({ma}_{14}left(Delta {n}_{W}left(tright)right)right),end{array}$$
(38)
where ({ma}_{14}left(cdot right)) is a centered moving average with window size of 14 days and ({rg}_{sigma }left(cdot right)) is a centered rolling average through a Gaussian window with standard deviation (sigma). The specific value of the window size has been chosen to mitigate weekly reporting effects. The values of the standard deviations of the Gaussian windows have been selected to achieve a smooth representation of the expected death estimation for each country as shown in the bottom panels of Supplementary Fig. S1.Reporting delaysWe consider an average delay of two days between reporting a death and its occurrence. This value is obtained by comparing the daily death counts reported for Spain1 and their actual values33 from February 15 to March 31, 2020. The values of the root-mean-squared deviation between reported and actual deaths shifted by 0, 1, 2, 3, and 4 days are 77.9, 58.4, 38.5, 58.7, and 88.6 deaths respectively.Infection fatality rate ((IFR))The infection fatality rate is computed assuming homogeneous attack rate as$$begin{array}{c}IFR=frac{1}{{sum }_{a}{g}_{a}}{sum }_{a}{IFR}_{a}{g}_{a} ,end{array}$$
(39)
where ({mathrm{IFR}}_{a}) is the previously estimated (IFR) for the age group (a)5 and ({g}_{a}) is the population in the age group (a) as reported by the United Nations for countries18 and the US Census for states19.Clinical parametersWe obtained the values of the average ({tau }_{G}) and standard deviation ({sigma }_{G}) of the generation time from Ref.13, of the averages of the incubation ({tau }_{I}) and symptom onset-to-death ({tau }_{OD}) times from Refs.5,14, and of the average number of days (Delta {t}_{TP}) of positive testing by an infected individual from Refs.15,17. The average time at which an individual tested positive after infection ({tau }_{TP}) was computed as ({tau }_{TP}={tau }_{I}-2+Delta {t}_{TP}/2), where we have assumed that on average an individual started to test positive 2 days before symptom onset. The average seroconversion time after infection ({tau }_{SP}) was estimated as ({tau }_{I}) plus the 7 days of 50% seroconversion after symptom onset reported in Ref.16.Dynamical constraints implementation with discrete timeWe implemented the dynamical constraints to compute the infectious and infected population as outlined in the main text and as detailed in the previous section of this document, using days as time units. Time delays were rounded to days to assign daily values.The first derivative of the cumulative number of deaths is computed as$$begin{array}{c}frac{d{n}_{D}left(tright)}{dt}=Delta {n}_{D}left(tright),end{array}$$
(40)
with (Delta {n}_{D}left(tright)={n}_{D}left(tright)-{n}_{D}(t-1)).The growth rate was computed explicitly from the discrete time series as the centered 7-day difference$$begin{array}{c}{k}_{G}left(tright)=frac{1}{7}left({mathrm{ln}}left(Delta {n}_{D}left(t+4right)+Delta {n}_{D}left(t+3right)right)-{mathrm{ln}}left(Delta {n}_{D}left(t-3right)+Delta {n}_{D}left(t-4right)right)right).end{array}$$
(41)
The 7-day value was chosen to mitigate reporting artifacts.Confidence and credibility intervalsConfidence intervals associated with death counts were computed using bootstrapping with 10,000 realizations34. These confidence intervals were combined with the credibility intervals of the (IFR) in infectious and infected populations assuming independence and additivity on a logarithmic scale.Fold accuracyThe fold accuracy, ({F}_{A}), is explicitly computed as$$begin{array}{c}{mathrm{log}}{F}_{A}=frac{1}{N}{sum }_{i=1}^{N}left|{mathrm{log}}{x}_{i}^{obs}-{mathrm{log}}{x}_{i}^{est}right|,end{array}$$
(42)
where (left|cdot right|) is the absolute value function, ({x}_{i}^{obs}) is the ({i}^{th}) observation, ({x}_{i}^{est}) is its corresponding estimation, and (N) is the total number of observations.Inference and extrapolationBecause of the delay between infections and deaths, inference for the values of the growth rate and infectious populations ends on December 30, 2020 and for the values of the infected populations ends on December 26, 2020. Extrapolation to the current time (January 21, 2021) is carried out assuming the last growth rate computed.Reproduction numberThe quantities ({R}_{t}) and ({k}_{G}left(tright)) are related to each other through the Euler–Lotka equation, ({R}_{t}^{-1}={int }_{0}^{infty }{f}_{GT}left(tau right){e}^{-{k}_{G}left(tright)tau }dtau ,) which considers (jleft(t-tau right)simeq {e}^{-{k}_{G}left(tright)tau }jleft(tright)) in the renewal equation (jleft(tright)={int }_{0}^{infty }{k}_{I}left(t,tau right)jleft(t-tau right)dtau). Generation times can generally be described through a gamma distribution ({f}_{GT}left(tau right)=frac{{beta }^{alpha }}{Gamma left(alpha right)}{tau }^{alpha -1}{e}^{-beta tau }) with (alpha ={tau }_{G}^{2}/{sigma }_{G}^{2}) and (beta ={tau }_{G}/{sigma }_{G}^{2}), which leads to ({R}_{t}={left(1+{k}_{G}(t)/beta right)}^{alpha }) for ({k}_{G}(t) >-beta) and ({R}_{t}=0) for ({k}_{G}left(tright)le -beta). In the case of the exponentially distributed limit ((alpha simeq 1)) or small values of ({k}_{G}(t)/beta), it simplifies to ({R}_{t}=1+{k}_{G}left(tright){tau }_{G}) for ({k}_{G}left(tright) >-1/{tau }_{G}) and ({R}_{t}=0) for ({k}_{G}left(tright)le -1/{tau }_{G}). Global prevalence data were obtained from multiple data sources35,36,37,38,39,40,41,42, as described in Supplementary Table S1. More

1.Voss, R. S. & Emmons, L. H. Mammalian diversity in Neotropical lowland rainforests: A preliminary assessment. Bull. Am. Museum Nat. Hist. 230, 1–115 (1996).
Google Scholar 
2.Barnosky, A. D. et al. Variable impact of late-Quaternary megafaunal extinction in causing ecological state shifts in North and South America. Proc. Natl. Acad. Sci. U. S. A. 113, 856–861 (2016).ADS 
CAS 
PubMed 
Article 

Google Scholar 
3.Croft, D. A., Engelman, R. K., Dolgushina, T. & Wesley, G. Diversity and disparity of sparassodonts (Metatheria) reveal non-analogue nature of ancient South American mammalian carnivore guilds. Proc. R. Soc. B 285, 20172012 (2018).PubMed 
PubMed Central 
Article 

Google Scholar 
4.Fariña, R. A. Trophic relationships among Lujanian mammals. Evol. Theory 11, 125–134 (1996).
Google Scholar 
5.Fariña, R. A. & Blanco, R. E. Megatherium the Stabber. Proc. R. Soc. B Biol. Sci. 263, 1725–1729 (2006).ADS 

Google Scholar 
6.Tejada-Lara, J. V. et al. Body mass predicts isotope enrichment in herbivorous mammals. Proc. R. Soc. B Biol. Sci. 285, 20181020 (2018).Article 
CAS 

Google Scholar 
7.de Muizon, C. & McDonald, H. G. An aquatic sloth from the Pliocene of Peru. Nature 375, 224–227 (1995).ADS 
Article 

Google Scholar 
8.Croft, D. A. The middle Miocene (Laventan) Quebrada Honda Fauna, southern Bolivia and a description of its notoungulates. Palaeontology 50, 277–303 (2007).Article 

Google Scholar 
9.Boecklen, W. J., Yarnes, C. T., Cook, B. A. & James, A. C. On the use of stable isotopes in trophic ecology. Annu. Rev. Ecol. Evol. Syst. 42, 411–440 (2011).Article 

Google Scholar 
10.Lee-Thorp, J. J., Sealy, J. J. C. & van der Merwe, N. J. N. Stable carbon isotope ratio differences between bone collagen and bone apatite, and their relationship to diet. J. Archaeol. Sci. 32, 1459–1470 (1989).
Google Scholar 
11.Clementz, M. T., Fox-Dobbs, K., Wheatley, P. V., Koch, P. L. & Doak, D. F. Revisiting old bones: Coupled carbon isotope analysis of bioapatite and collagen as an ecological and palaeoecological tool. Geol. J. 44, 605–620 (2009).CAS 
Article 

Google Scholar 
12.Tejada, J. V. et al. Comparative isotope ecology of western Amazonian rainforest mammals. Proc. Natl. Acad. Sci. https://doi.org/10.1073/pnas.2007440117 (2020).Article 
PubMed 
PubMed Central 

Google Scholar 
13.Robinson, D. δ15N as an integrator of the nitrogen cycle. Trends Ecol. Evol. 16, 153–162 (2001).CAS 
PubMed 
Article 

Google Scholar 
14.McMahon, K. W. & McCarthy, M. D. Embracing variability in amino acid δ15N fractionation: Mechanisms, implications, and applications for trophic ecology. Ecosphere 7, 1–26 (2016).Article 

Google Scholar 
15.McClelland, J. W. & Montoya, J. P. Trophic relationships and the nitrogen isotopic composition of amino acids in plankton. Ecology 83, 2173–2180 (2002).Article 

Google Scholar 
16.Chikaraishi, Y., Ogawa, N. O., Doi, H. & Ohkouchi, N. 15N/14N ratios of amino acids as a tool for studying terrestrial food webs: A case study of terrestrial insects (bees, wasps, and hornets ). Ecol. Res. 26, 835–844 (2011).Article 

Google Scholar 
17.Popp, B. N. et al. Insight into the trophic ecology of yellowfin tuna, Thunnus albacares, from compound- specific nitrogen isotope analysis of proteinaceous amino acids. In Stable Isotopes as Indicators of Ecological Change (eds Dawson, T. E. & Siegwolf, R. T. W.) 173–190 (Elsevier Inc., 2007).
Google Scholar 
18.Naito, Y. I., Honch, N. V., Chikaraishi, Y., Ohkouchi, N. & Yoneda, M. Quantitative evaluation of marine protein contribution in ancient diets based on nitrogen isotope ratios of individual amino acids in bone collagen: An investigation at the Kitakogane Jomon Site. Am. J. Phys. Anthropol. 143, 31–40 (2010).PubMed 
Article 

Google Scholar 
19.O’Connell, T. C. ‘Trophic’ and ‘source’ amino acids in trophic estimation: A likely metabolic explanation. Oecologia 184, 317–326 (2017).ADS 
PubMed 
PubMed Central 
Article 

Google Scholar 
20.Chikaraishi, Y., Ogawa, N. O. & Ohkouchi, N. Further evaluation of the trophic level estimation based on nitrogen isotopic composition of amino acids. In Earth, Life, and Isotopes (eds Ohkouchi, N. et al.) 37–51 (Kyoto Universy Press, 2010).
Google Scholar 
21.Steffan, S. A. et al. Trophic hierarchies illuminated via amino acid isotopic analysis. PLoS ONE 8, 1–10 (2013).Article 
CAS 

Google Scholar 
22.Chikaraishi, Y., Kashiyama, Y., Ogawa, N. O., Kitazato, H. & Ohkouchi, N. Metabolic control of nitrogen isotope composition of amino acids in macroalgae and gastropods: Implications for aquatic food web studies. Mar. Ecol. Prog. Ser. 342, 85–90 (2007).ADS 
CAS 
Article 

Google Scholar 
23.Naito, Y. I. et al. Ecological niche of Neanderthals from Spy Cave revealed by nitrogen isotopes of individual amino acids in collagen. J. Hum. Evol. 93, 82–90 (2016).PubMed 
Article 

Google Scholar 
24.Nielsen, J. M., Popp, B. N. & Winder, M. Meta-analysis of amino acid stable nitrogen isotope ratios for estimating trophic position in marine organisms. Oecologia https://doi.org/10.1007/s00442-015-3305-7 (2015).Article 
PubMed 

Google Scholar 
25.Décima, M., Landry, M. R. & Popp, B. N. Environmental perturbation effects on baseline δ15N values and zooplankton trophic flexibility in the southern California current ecosystem. Limnol. Oceanogr. 58, 624–634 (2013).ADS 
Article 
CAS 

Google Scholar 
26.Jarman, C. L. et al. Diet of the prehistoric population of Rapa Nui (Easter Island, Chile) shows environmental adaptation and resilience. Am. J. Phys. Anthropol. https://doi.org/10.1002/ajpa.23273 (2017).Article 
PubMed 
PubMed Central 

Google Scholar 
27.Kendall, I. P. et al. Compound-specific δ15N values express differences in amino acid metabolism in plants of varying lignin content. Phytochemistry 161, 130–138 (2019).CAS 
PubMed 
Article 

Google Scholar 
28.Ramirez, M. D., Besser, A. C., Newsome, S. D. & McMahon, K. W. Meta-analysis of primary producer amino acid δ15N values and their influence on trophic position estimation. Methods Ecol. Evol. https://doi.org/10.1111/2041-210X.13678 (2021).Article 

Google Scholar 
29.Hebert, C. E., Popp, B. N., Fernie, K. J., Rattner, B. A. & Wallsgrove, N. Amino acid specific stable nitrogen isotope values in avian tissues: Insights from captive American kestrels and wild herring gulls. Environ. Sci. Technol. 50, 12928–12937 (2016).ADS 
CAS 
PubMed 
Article 

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

Google Scholar 
31.Steffan, S. A. et al. Microbes are trophic analogs of animals. Proc. Natl. Acad. Sci. 112, 15119–15124 (2015).ADS 
CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
32.Kendall, I. P., Lee, M. R. F. & Evershed, R. P. The effect of trophic level on individual amino acid δ15N values in a terrestrial ruminant food web. Sci. Technol. Archaeol. Res. 3, 135–145 (2017).
Google Scholar 
33.Matthews, C. J. D., Ruiz-Cooley, R. I., Pomerleau, C. & Ferguson, S. H. Amino acid δ15N underestimation of cetacean trophic positions highlights limited understanding of isotopic fractionation in higher marine consumers. Ecol. Evol. 10, 3450–3462 (2020).PubMed 
PubMed Central 
Article 

Google Scholar 
34.Styring, A. K., Sealy, J. C. & Evershed, R. P. Resolving the bulk δ15N values of ancient human and animal bone collagen via compound-specific nitrogen isotope analysis of constituent amino acids. Geochim. Cosmochim. Acta 74, 241–251 (2010).ADS 
CAS 
Article 

Google Scholar 
35.Lorrain, A. et al. Nitrogen and carbon isotope values of individual amino acids: A tool to study foraging ecology of penguins in the Southern Ocean. Mar. Ecol. Prog. Ser. 391, 293–306 (2009).ADS 
CAS 
Article 

Google Scholar 
36.Lorrain, A. et al. Nitrogen isotopic baselines and implications for estimating foraging habitat and trophic position of yellowfin tuna in the Indian and Pacific Oceans. Deep. Res. Part II Top. Stud. Oceanogr. 113, 188–198 (2015).ADS 
CAS 
Article 

Google Scholar 
37.Hartman, G. Are elevated δ15N values in herbivores in hot and arid environments caused by diet or animal physiology?. Funct. Ecol. 25, 122–131 (2011).Article 

Google Scholar 
38.Hartman, G. & Danin, A. Isotopic values of plants in relation to water availability in the Eastern Mediterranean region. Oecologia 162, 837–852 (2010).ADS 
PubMed 
Article 
PubMed Central 

Google Scholar 
39.Hansen, R. M. Shasta ground sloth food habits, Rampart Cave, Arizona. Paleobiology 4, 302–319 (1978).Article 

Google Scholar 
40.McDonald, H. G. & Morgan, G. S. Ground Sloths of New Mexico. Foss. Rec. 3 New. Mex. Museum Nat. Hist. Sci. Bull. 53, 652–663 (2011).
Google Scholar 
41.Poinar, H. N. Molecular coproscopy: Dung and diet of the extinct ground sloth Nothrotheriops shastensis. Science 281, 402–406 (1998).ADS 
CAS 
PubMed 
Article 

Google Scholar 
42.Clack, A. A., MacPhee, R. D. E. & Poinar, H. N. Mylodon darwinii DNA sequences from ancient fecal hair shafts. Ann. Anat. 194, 26–30 (2012).CAS 
PubMed 
Article 

Google Scholar 
43.Höss, M., Dilling, A., Currant, A. & Pääbo, S. Molecular phylogeny of the extinct ground sloth Mylodon darwinii. Proc. Natl. Acad. Sci. U. S. A. 93, 181–185 (1996).ADS 
PubMed 
PubMed Central 
Article 

Google Scholar 
44.Moore, D. M. Post-glacial vegetation in the South Patagonian territory of the giant ground sloth, Mylodon. Bot. J. Linn. Soc. 77, 177–202 (1978).Article 

Google Scholar 
45.Bargo, M. S., Toledo, N. & Vizcaino, S. F. Muzzle of South American Pleistocene ground sloths (Xenarthra, Tardigrada). J. Morphol. 267, 248–263 (2006).PubMed 
Article 

Google Scholar 
46.Rasmussen, M. et al. Response to comment by Goldberg et al. on ‘DNA from Pre-Clovis human coprolites in Oregon, North America’. Science 325, 148 (2009).ADS 
CAS 
Article 

Google Scholar 
47.Janis, C. M. Correlations between craniodental anatomy and feeding in ungulates: Reciprocal illumination between living and fossil taxa. In Functional Morphology in Vertebrate Paleontology (ed. Thomason, J.) 76–98 (Cambridge U Press, 1995).
Google Scholar 
48.Clauss, M., Nunn, C., Fritz, J. & Hummel, J. Evidence for a tradeoff between retention time and chewing efficiency in large mammalian herbivores. Comp. Biochem. Physiol. A Mol. Integr. Physiol. 154, 376–382 (2009).PubMed 
Article 
CAS 

Google Scholar 
49.Vizcaino, S. F., Bargo, M. S. & Cassini, G. H. Dental occlusal surface area in relation to body mass, food habits and other biologic features in fossil xenarthrans. Ameghiniana 43, 11–26 (2006).
Google Scholar 
50.McNab, B. K. Energetics, population biology, and distribution of xenarthrans, living and extinct. In The Ecology of Arboreal Folivores 219–232 (Smithsonian Press, 1985).51.Davis, L. B. & Birkbak, R. C. On the transfer of energy in layers of fur. Biophys. J. 14, 249–268 (1974).PubMed 
PubMed Central 
Article 

Google Scholar 
52.Clauss, M. et al. The maximum attainable body size of herbivorous mammals: Morphophysiological constraints on foregut, and adaptations of hindgut fermenters. Oecologia 136, 14–27 (2003).ADS 
CAS 
PubMed 
Article 

Google Scholar 
53.Fariña, R. A., Czerwonogora, A. & Di Giacomo, M. Splendid oddness: Revisiting the curious trophic relationships of South American Pleistocene mammals and their abundance. An. Acad. Bras. Cienc. 86, 311–331 (2014).PubMed 
Article 

Google Scholar 
54.Zhu, D. et al. The large mean body size of mammalian herbivores explains the productivity paradox during the Last Glacial Maximum. Nat. Ecol. Evol. 2, 640–649 (2018).PubMed 
PubMed Central 
Article 

Google Scholar 
55.Zimov, S. A., Zimov, N. S., Tikhonov, A. N. & Chapin, I. S. Mammoth steppe: A high-productivity phenomenon. Quat. Sci. Rev. 57, 26–45 (2012).ADS 
Article 

Google Scholar 
56.Hannides, C. C. S., Popp, B. N., Landry, M. R. & Graham, B. S. Quantification of zooplankton trophic position in the North Pacific Subtropical Gyre using stable nitrogen isotopes. Limnol. Oceanogr. 54, 50–61 (2009).ADS 
CAS 
Article 

Google Scholar  More