Ecology

Our findings indicated that mine water had a certain inhibitory effect on ryegrass seed germination, and the intensity of this inhibitory effect increased with increased mine water proportion. These effects were mainly reflected as changes in germination potential. Concretely, irrigation with mine water prolonged the germination of ryegrass seeds but had no significant effect on germination rate. Min Zhu et al. found that recycled water inhibited the seed germination of turfgrass, and this effect became more notorious when the concentration of reclaimed water increased. This was likely because the water contained salt ions, heavy metal ions, and E. coli, all of which are known to affect seed germination22. The mine water was taken from the Laohutai mining area, where the water composition and quality are good. Therefore, mine water did not significantly affect seed germination and the seeds maybe germinate normally if given sufficient time.The physiological and photosynthetic characteristics of ryegrass were impacted by the mine water, and the intensity of inhibition increased with higher mine water proportions. When the ratio of mine water to clean water reached a certain proportion (1:2, A1 and B1), the physiological and growth characteristics of ryegrass were improved to a certain extent. When only mine water was used for irrigation, the indices were significantly suppressed. In contrast, mixing clear water with mine water for irrigation promoted the physiological characteristics of the plants, as well as photosynthesis. This was likely because the mineral content of mine water is higher. However, mine water not only contains elements needed for plant growth but also some elements and ions that have inhibitory effects on plant growth. Therefore, the quality of ryegrass growth were suppressed when irrigating only with mine water. In contrast, after mixing the mine water with clean water, the concentrations of certain substances that produce adverse effects are diluted, and the mixed irrigation water promoted ryegrass growth in appropriate proportion.A certain concentration of heavy metal elements will affect the absorption of essential elements by plants and produce antagonism, and high concentration can directly lead to plant death. Heavy metal stress affects chlorophyll content through two aspects: Heavy metal destroys enzymes needed for chlorophyll synthesis, affects plant chlorophyll synthesis, and then inhibits plant photosynthesis23. The second is the destruction of chloroplast structure and cell membrane24,25,26. In the treatment of high concentrations of heavy metals, the chlorophyll content of plants is significantly reduced due to the inhibition of chlorophylase or aminolevulinic acid dehydrase, thus inhibiting plant photosynthesis27. The heavy metal threat forcing stimulates the formation of reactive oxygen species that convert fatty acids into toxic lipid peroxides, which damage to plant cells28,29,30. Heavy metal stress can induce a lot of activity in plants sexual oxygen and inhibit the normal metabolism of plants, causing membrane lipid peroxidation and increased plasma membrane permeability31, 32. Low concentration of heavy metal stress will stimulate the protective mechanism of plants, so low concentration of stimulation will not damage plants, on the contrary, may help plant growth. Heavy metal stress causes water loss in plants, and a certain amount of proline can be produced to regulate the water balance of plant cells and reduce the damage degree of plant cells33. SOD, POD and CATT are important antioxidant enzymes in plants, which can scavenge excessive free radicals. The synergistic action of three enzymes can protect plants from free radical damage. When the concentration of heavy metals was low, the activity of protective enzymes increased under the induction of reactive oxygen radicals. However, with the increase of stress degree, the activities of SOD, POD and CAT decreased, which eventually led to the persecution of plant cells34. These conclusions are consistent with the results of this paper. When mine water was mixed with clear water at a ratio of 1:2, heavy metal stress stimulated the protective mechanism of ryegrass most appropriately, and improved plant quality and resistance. On the contrary, when the proportion of mine water increased, the physiological characteristics and quality of ryegrass plants were inhibited to different degrees.Precious Nneka Amori et al. studied physiological traits of leaves and Silverbeet using treated wastewater, the results show that the biomass of plants watered with only the treated wastewater were more than 50% higher than the yield in tap water control and plants exhibited high degree of root foraging1. Libutti et al. irrigated tomato and broccoli with purified agro-industrial effluent and reported that yield and quality traits of agricultural products were not affected35. Radish was grown using a reclaimed synthetic textile wastewater treated in an anoxic-aerobic photobioreactor, and the dry weight, leaf number and leaf area of plant harvest were 49, 19.2 and 62% higher than the growth performance in freshwater irrigation36. FU et al. studied four native Chenopodiaceae plants of Halogeton glomeratus, Kochia scoparia, Suaeda glauca and Chenopodium glaucum in Jinchang area northwest China, from their changes of net photosynthetic rate (Pn), Stomatal conductance (Gs), transpiration rate (Tr). chlorophyll content (Chl), malondialdehyde (MDA), soluble protein (SP), proline (Pro) and antioxidant enzymes activity under the treatment of farmland soil (T1) and sedendary soil mixed with tailing (1:1, T2), they concluded that under T2 treatment, Pn, Gs, and Tr of Halogeton glomeratus and Kochia scoparia were decreased , the other six indexes were increased significantly. Gs, Tr, MDA, Pro, and SOD increased, yet CAT, Chl and Pn of Suaeda glauca decreased significantly, respectively. Pn, Gs, and Tr of Chenopodium glaucum decreased significantly, while SP, POD increased significantly37. Our results also indicated that mine water irrigation had significant effects on soil characteristics. At higher mine water ratios, the soil conductivity increased exponentially, the pH decreased gradually, the content of K+, Na+, Ca2+ and Mg2+ increased, and the content of N, P and K also increased gradually. In contrast, the clean water and mine water mixture had little effect on the soil properties. This was because the salt and metal ions in mine water migrate to the soil during the irrigation process, which significantly changes the soil properties. As a result, the concentration of salt in the soil increased and soil acidity also increased. After mixing with clean water, the concentration of salt decreases, and the influence on the soil matrix weakened. These results also indicated that the growth, physiological, and photosynthetic effects of ryegrass in the pot experiments were better than those in soilless culture, because there were many other organic materials and inorganic ions in soil that could promote growth, whereas the plants in the hydroponic system lacked other nutrients that benefit plant growth. Many existing studies have shown that mine drainage or other wastewater can improve the quality and yield of one or more kinds of plants to different degrees after certain treatment, but some studies also show that the reclaimed water used for irrigation will cause harm to plants, soil and even human health.Jinfang Yang et al. reported that long-term irrigation with mine water significantly reduced the soil respiration rate and soil enzyme activity. Mine water irrigation also significantly inhibited wheat plant height, leaf area, chlorophyll content, and photosynthetic rate, and wheat production was also markedly reduced38. Jianjun Cha found that acidic mining waste water can reduce the pH of the soil profile and increase its electrical conductivity39. Junhao Qin et al. found that if treated mine water is used as an irrigation water source, acidic substances may still be introduced into the soil. This inhibits plant growth and may also enhance leaching of some trace elements in the soil to shallow aquifers, resulting in groundwater pollution40, 41. The results of this study are consistent with the above conclusions, that is, directly irrigating with mine water can significantly inhibit plant growth and photosynthesis, thus affecting the quality of ryegrass plants. MA et al. studied the effects of irrigation with mine wastewater on the physiological characters and heavy metals accumulation of winter wheat. It shows that irrigation with mine wastewater had negative effects on the winter wheat growth and grain yield. At anthesis stage, the leaf area, dry mass per stem, root activity and net photosynthetic rate of winter wheat in treatments were significantly lower and the plant height and leaf chlorophyll content was decreased. In addition, the heavy metals (Cr, Pb, Cu and Zn) contents in the grain of winter wheat under mine wastewater irrigation were significantly higher than those in control, it suggested that the irrigation with mine water could result in the heavy metals accumulation in wheat grain42. A large number of studies have shown that direct use of mine water for irrigation will have a negative impact on soil and plants, but this study found that after a certain processing of mine water (mine water was mixed with clear water in a ratio of 1:2) used for irrigation does not significantly alter soil properties, but can increase plant yield and quality, it will be meaningful to mine water reuse, soil utilization around the mining area and the agriculture.The conclusions of this study are based on mine water from Fushun mining area in Northeast China, but the effects of mine water on plants from other mining areas are uncertain. At the same time, ryegrass, a cold-season turfgrass, is only selected in this study. If it is other kinds of plants, how they respond to mine water irrigation needs further study. What are the effects of mine water irrigation on plants other than ryegrass that need further study. Moreover, this study was only a short-term experiment, and the effects of mine water on the properties of the soil matrix cannot be generalized. Indoor experiments can be regularly watered to maintain moisture, indoor temperature is relatively fixed, while the natural environment is a lot of uncertainty and uncontrollable. Would the results of a small-scale pot experiment in a controlled environment be different if it were applied to a field where there are many uncertainties about soil properties and atmospheric conditions? Long-term field experiments must also be conducted to confirm our findings in more realistic conditions. The use of mine water resources not only has environmental and social benefits but could also bring economic benefits43. This study demonstrated that mine water can be used in ecological restoration and agricultural irrigation in mining areas, and is therefore of great significance to environmental restoration. More

Ceballos, G. et al. Accelerated modern human–induced species losses: Entering the sixth mass extinction. Sci. Adv. 1, e1400253. https://doi.org/10.1126/sciadv.1400253 (2015).Article 
ADS 
PubMed 
PubMed Central 

Google Scholar 
Bradshaw, C. J. A. et al. Underestimating the challenges of avoiding a ghastly future. Front. Environ. Sci. 1, 615469. https://doi.org/10.3389/fcosc.2020.615419 (2021).Article 

Google Scholar 
IUCN. The IUCN Red List of Threatened Species. IUCN, Gland, Switzerland. http://www.iucnredlist.org (2020).Murray, K. A., Verde Arregoitia, L. D., Davidson, A., Di Marco, M. & Di Fonzo, M. M. I. Threat to the point: Improving the value of comparative extinction risk analysis for conservation action. Glob. Chang. Biol. 20, 483–494 (2014).Article 
ADS 

Google Scholar 
Benítez-López, A. et al. The impact of hunting on tropical mammal and bird populations. Science 356, 180–183 (2017).Article 
ADS 

Google Scholar 
Chichorro, F., Juslén, A. & Cardoso, P. A review of the relation between species traits and extinction risk. Biol. Conserv. 237, 220–229 (2019).Article 

Google Scholar 
Ripple, W. J. et al. Bushmeat hunting and extinction risk to the world’s mammals. R. Soc. Open. Sci. 3, 160498. https://doi.org/10.1098/rsos.160498 (2016).Article 
ADS 
PubMed 
PubMed Central 

Google Scholar 
Hoffmann, M. et al. The changing fates of the world’s mammals. Phil. Trans. R. Soc. B. 366, 2598–2610 (2011).Article 

Google Scholar 
Verde Arregoitia, L. D. Biases, gaps, and opportunities in mammalian extinction risk research. Mammal. Rev. 46, 17–29 (2016).Article 

Google Scholar 
Di Marco, M. et al. Drivers of extinction risk in African mammals: The interplay of distribution state, human pressure, conservation response and species biology. Philos. Trans. R. Soc. Lond. B. 369, 1–12 (2014).Article 

Google Scholar 
Di Marco, M., Collen, B., Rondinini, C. & Mace, G. M. Historical drivers of extinction risk: Using past evidence to direct future monitoring. Proc. R. Soc. B. 282, 20150928. https://doi.org/10.1098/rspb.2015.0928 (2015).Article 
PubMed 
PubMed Central 

Google Scholar 
Bogoni, J. A., Ferraz, K. M. & Peres, C. A. Continental-scale local extinctions in mammal assemblages are synergistically induced by habitat loss and hunting pressure. Biol. Conserv. 272, 109635. https://doi.org/10.1016/j.biocon.2022.109635 (2022).Article 

Google Scholar 
Yusefi, G. H., Faizolahi, K., Darvish, J., Safi, K. & Brito, J. C. The species diversity, distribution and conservation status of the terrestrial mammals of Iran. J. Mammal. 100, 55–71 (2019).Article 

Google Scholar 
Keil, P. et al. Spatial scaling of extinction rates: Theory and data reveal nonlinearity and a major upscaling and downscaling challenge. Glob. Ecol. Biogeogr. 27, 2–13 (2018).Article 

Google Scholar 
Howard, C., Flather, C. H. & Stephens, P. A. A global assessment of the drivers of threatened terrestrial species richness. Nat. Commun. 11, 993. https://doi.org/10.1038/s41467-020-14771-6 (2020).Article 
ADS 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Rodríguez, J. P. The difference conservation can make: integrating knowledge to reduce extinction risk. Oryx 51, 1–2 (2017).Article 

Google Scholar 
Cardillo, M. & Meijaard, E. Are comparative studies of extinction risk useful for conservation?. Trends Ecol. Evol. 27, 167–171 (2012).Article 

Google Scholar 
Davidson, A. D. et al. Geography of current and future global mammal extinction risk. PLoS ONE 12, e0186934. https://doi.org/10.1371/journal.pone.0186934 (2017).Article 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Collen, B., Bykova, E., Ling, S., Milner-Gulland, E. J. & Purvis, A. Extinction risk: A comparative analysis of central Asian vertebrates. Biodivers. Conserv. 15, 1859–1871 (2006).Article 

Google Scholar 
Peñaranda, D. A. & Simonetti, J. A. Predicting and setting conservation priorities for Bolivian mammals based on biological correlates of the risk of decline. Conserv. Biol. 29, 834–843 (2015).Article 

Google Scholar 
Fritz, S. A., Bininda-Emonds, O. R. & Purvis, A. Geographical variation in predictors of mammalian extinction risk: Big is bad, but only in the tropics. Ecol. Lett. 12, 538–549 (2009).Article 

Google Scholar 
Cardillo, M. et al. Human population density and extinction risk in the world’s carnivores. PLoS Biol. 2, 909–914 (2004).Article 
CAS 

Google Scholar 
Cardillo, M. et al. The predictability of extinction: Biological and external correlates of decline in mammals. Proc. R. Soc. B. 275, 1441–1448 (2008).Article 

Google Scholar 
Yackulic, C. B., Sanderson, E. W. & Uriat, M. Anthropogenic and environmental drivers of modern range loss in large mammals. Proc. Natl. Acad. Sci. USA 108, 4024–4029 (2011).Article 
ADS 
CAS 

Google Scholar 
Ripple, W. J. et al. Are we eating the world’s megafauna to extinction?. Conserv Lett 12, e12627. https://doi.org/10.1111/conl.12627 (2019).Article 

Google Scholar 
Ripple, W. J. et al. Extinction risk is most acute for the world’s largest and smallest vertebrates. PNAS https://doi.org/10.1073/pnas.1702078114 (2017).Article 
PubMed 
PubMed Central 

Google Scholar 
Bodmer, R. E., Eisenberg, J. E. & Redford, K. H. Hunting and the likelihood of extinction of Amazonian mammals. Conserv. Biol. 11, 460–466 (1997).Article 

Google Scholar 
Lee, T. M. & Jetz, W. Unravelling the structure of species extinction risk for predictive conservation science. Proc. R. Soc. B. 278, 1329–1338 (2011).Article 

Google Scholar 
Wolf, C. & Ripple, W. J. Prey depletion as a threat to the world’s large carnivores. Roy. Soc Open Sci 3, 160252. https://doi.org/10.1098/rsos.160252 (2016).Article 
ADS 

Google Scholar 
Firouz, E. The complete fauna of Iran. I. B. (Tauris and Co Ltd, London, 2005).Cardillo, M. et al. Multiple causes of high extinction risk in large mammal species. Science 309, 1239–1241 (2005).Article 
ADS 
CAS 

Google Scholar 
Davidson, A. D., Hamilton, M. J., Boyer, A. G., Brown, J. H. & Ceballos, G. Multiple ecological pathways to extinction in mammals. Proc. Natl. Acad. Sci. USA 106, 10702–10705 (2009).Article 
ADS 
CAS 

Google Scholar 
Hill, J., DeVault, T. & Belant, J. Comparative influence of anthropogenic landscape pressures on cause-specific mortality of mammals. Perspect. Ecol. Conserv. 20, 38–44 (2022).
Google Scholar 
DOE-GIS. Areas under protection by the Department of Environment of Iran. Department of the Environment of Iran: GIS and Remote Sensing Section (2016).Kolahi, M., Sakai, T., Moriya, K. & Makhdoum, M. F. Challenges to the future development of Iran’s protected areas system. Environ. Manage. 50, 750–765 (2012).Article 
ADS 

Google Scholar 
Morrison, J. M., Sechrest, W., Dinerstein, E., Wilcove, D. S. & Lamoreux, J. L. Persistence of large mammal faunas as indicators of human impact. J. Mammal. 88, 1363–1380 (2007).Article 

Google Scholar 
Ghoddousi, A. et al. The decline of ungulate populations in Iranian protected areas calls for urgent action against poaching. Oryx 53, 151–158 (2017).Article 

Google Scholar 
Soofi, M. et al. Assessing the relationship between illegal hunting of ungulates, wild prey occurrence and livestock depredation rate by large carnivores. J. Appl. Ecol. 56, 365–374 (2019).Article 

Google Scholar 
Khalatbari, L., Yusefi, G. H., Martinez-Freiria, F., Jowkar, H. & Brito, J. C. Availability of prey and natural habitats are related with temporal dynamics in range and habitat suitability for Asiatic cheetah. Hystrix Ital. J. Mammal. 29, 145–151 (2018).
Google Scholar 
Ripple, W. J. et al. Collapse of the world’s largest herbivores. Sci. Adv. 1, e1400103 (2015).Article 
ADS 

Google Scholar 
Hoffmann, M. et al. The difference conservation makes to extinction risk of the world’s ungulates. Conserv. Biol. 29, 1303–1313 (2015).Article 

Google Scholar 
Yusefi, G. H. Conservation biogeography of terrestrial mammals in Iran diversity distribution and vulnerability to extinction. Front Biogeogr 13(2), 49765. https://doi.org/10.21425/F5FBG49765 (2021).Article 

Google Scholar 
Faurby, S. & Svenning, J.-C. A species-level phylogeny of all extant and late Quaternary extinct mammals using a novel heuristic-hierarchical Bayesian approach. Mol. Phylogenet. Evol. 84, 14–26 (2015).Article 

Google Scholar 
R Development Core Team. R: a language and environment for statistical computing. R Foundation for Statistical Computing (2021).Paradis, E. & Schliep, K. ape 5.0: An environment for modern phylogenetics and evolutionary analyses in R. Bioinformatics 35, 526–528 (2018).Article 

Google Scholar 
Jones, K. et al. PanTHERIA: A species-level database of life history, ecology, and geography of extant and recently extinct mammals. Ecology 90, 2648. https://doi.org/10.1890/08-1494.1 (2009).Article 

Google Scholar 
González-Suárez, M., Lucas, P. M. & Revilla, E. Biases in comparative analyses of extinction risk: Mind the gap. J. Anim. Ecol. 81, 1211–1222 (2012).Article 

Google Scholar 
Wang, Y. et al. Ecological correlates of extinction risk in Chinese birds. Ecography 41, 782–794 (2018).Article 

Google Scholar 
Wildlife Conservation Society-WCS, and Center for International Earth Science Information Network-CIESIN, Columbia University. Last of the wild project, Version 2, 2005 (LWP-2): Global human influence index (HII) Dataset. https://sedac.ciesin.columbia.edu/data/set/wildareas-v2-human-influence-index-geographic (2005).ESRI ArcGIS Desktop10.6. Redlands, CA: Environmental Systems Research Institute (2017).Hijmans, R. J. et al. raster: geographic data analysis and modeling. https://cran.r-project.org/web/packages/raster/index.html (2018).Pebesma, E. et al. rgdal: bindings for the geospatial data abstraction library. https://cran.r-project.org/web/packages/rgdal/index.html (2018).Bivand, R. et al. maptools: tools for reading and handling spatial objects. https://cran.r-project.org/web/packages/maptools/ index.html (2018).Purvis, A. Phylogenetic approaches to the study of extinction. Annu. Rev. Ecol. Evol. Syst. 39, 301–319 (2008).Article 

Google Scholar 
Venables, W. N. & Ripley, B. D. Modern applied statistics with S (Springer, 2002).Book 

Google Scholar 
Gittleman, J. L. & Kot, M. Adaptation: Statistics and a null model for estimating phylogenetic effects. Syst. Zool. 39, 227–241 (1990).Article 

Google Scholar 
Zuur, A. F., Ieno, E. N. & Elphick, C. S. A protocol for data exploration to avoid common statistical problems. Methods Ecol. Evol. 1, 3–14 (2010).Article 

Google Scholar 
Grueber, C. E., Nakagawa, S., Laws, R. J. & Jamieson, I. G. Multimodal inference in ecology and solution: Challenges and solutions. J. Evol. Biol. 24, 699–711 (2011).Article 
CAS 

Google Scholar 
Bartón, K. MuMIn: multi-model inference R package version 1.43.17. https://CRAN.R-project.org/package=MuMIn (2020). More

Peñuelas, J. et al. Assessment of the impacts of climate change on mediterranean terrestrial ecosystems based on data from field experiments and long-term monitored field gradients in Catalonia. Environ. Exp. Bot. 152, 49–59 (2018).
Google Scholar 
Pugnaire, F. I. et al. Climate change effects on plant-soil feedbacks and consequences for biodiversity and functioning of terrestrial ecosystems. Sci. Adv. 5, eaaz1834 (2019).ADS 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Baho, D. L., Bundschuh, M. & Futter, M. N. Microplastics in terrestrial ecosystems: Moving beyond the state of the art to minimize the risk of ecological surprise. Glob. Change Biol. 27, 3969–3986 (2021).CAS 

Google Scholar 
Rillig, M. C., Kim, S. W., Kim, T.-Y. & Waldman, W. R. The global plastic toxicity debt. Environ. Sci. Technol. 55, 2717–2719 (2021).ADS 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Nizzetto, L., Futter, M. & Langaas, S. Are agricultural soils dumps for microplastics of urban origin?. (2016).Geyer, R., Jambeck, J. R. & Law, K. L. Production, use, and fate of all plastics ever made. Sci. Adv. 3, e1700782 (2017).ADS 
PubMed 
PubMed Central 

Google Scholar 
Barnes, D. K., Galgani, F., Thompson, R. C. & Barlaz, M. Accumulation and fragmentation of plastic debris in global environments. Philos. Trans. R. Soc. B Biol. Sci. 364, 1985–1998 (2009).CAS 

Google Scholar 
Rillig, M. C. Microplastic in terrestrial ecosystems and the soil?. (2012).de Souza Machado, A. A. et al. Microplastics can change soil properties and affect plant performance. Environ. Sci. Technol. 53, 6044–6052 (2019).ADS 
PubMed 

Google Scholar 
Büks, F. & Kaupenjohann, M. Global concentrations of microplastics in soils–A review. Soil 6, 649–662 (2020).ADS 

Google Scholar 
Evangeliou, N. et al. Atmospheric transport is a major pathway of microplastics to remote regions. Nat. Commun. 11, 1–11 (2020).
Google Scholar 
Steinmetz, Z. et al. Plastic mulching in agriculture. Trading short-term agronomic benefits for long-term soil degradation?. Sci. Total Environ. 550, 690–705 (2016).ADS 
CAS 
PubMed 

Google Scholar 
de Souza Machado, A. A., Kloas, W., Zarfl, C., Hempel, S. & Rillig, M. C. Microplastics as an emerging threat to terrestrial ecosystems. Glob. Change Biol. 24, 1405–1416 (2018).ADS 

Google Scholar 
Qi, Y. et al. Macro-and micro-plastics in soil-plant system: Effects of plastic mulch film residues on wheat (Triticum aestivum) growth. Sci. Total Environ. 645, 1048–1056 (2018).ADS 
CAS 
PubMed 

Google Scholar 
Weithmann, N. et al. Organic fertilizer as a vehicle for the entry of microplastic into the environment. Sci. Adv. 4, eaap8060 (2018).ADS 
PubMed 
PubMed Central 

Google Scholar 
Allen, S. et al. Atmospheric transport and deposition of microplastics in a remote mountain catchment. Nat. Geosci. 12, 339–344 (2019).ADS 
CAS 

Google Scholar 
Kiyama, Y., Miyahara, K. & Ohshima, Y. Active uptake of artificial particles in the nematode Caenorhabditis elegans. J. Exp. Biol. 215, 1178–1183 (2012).PubMed 

Google Scholar 
Helmberger, M. S., Tiemann, L. K. & Grieshop, M. J. Towards an ecology of soil microplastics. Funct. Ecol. 34, 550–560 (2020).
Google Scholar 
Liu, M. et al. Microplastic and mesoplastic pollution in farmland soils in suburbs of Shanghai, China. Environ. Pollut. 242, 855–862 (2018).CAS 
PubMed 

Google Scholar 
Lehmann, A., Fitschen, K. & Rillig, M. C. Abiotic and biotic factors influencing the effect of microplastic on soil aggregation. Soil Syst. 3, 21 (2019).CAS 

Google Scholar 
de Souza Machado, A. A. et al. Impacts of microplastics on the soil biophysical environment. Environ. Sci. Technol. 52, 9656–9665 (2018).ADS 
PubMed 
PubMed Central 

Google Scholar 
Kim, S. W. & Rillig, M. C. Research trends of microplastics in the soil environment: Comprehensive screening of effects. Soil Ecol. Lett. 4, 109–118 (2022).CAS 

Google Scholar 
Lehmann, A., Leifheit, E. F., Gerdawischke, M. & Rillig, M. C. Microplastics have shape-and polymer-dependent effects on soil aggregation and organic matter loss–An experimental and meta-analytical approach. Microplast. Nanoplast. 1, 1–14 (2021).
Google Scholar 
Wan, Y., Wu, C., Xue, Q. & Hui, X. Effects of plastic contamination on water evaporation and desiccation cracking in soil. Sci. Total Environ. 654, 576–582 (2019).ADS 
CAS 
PubMed 

Google Scholar 
Liang, Y., Lehmann, A., Yang, G., Leifheit, E. F. & Rillig, M. C. Effects of microplastic fibers on soil aggregation and enzyme activities are organic matter dependent. Front. Environ. Sci. 9, 97 (2021).
Google Scholar 
Lozano, Y. M., Lehnert, T., Linck, L. T., Lehmann, A. & Rillig, M. C. Microplastic shape, polymer type, and concentration affect soil properties and plant biomass. Front. Plant Sci. 12, 169 (2021).
Google Scholar 
Kemper, W. Aggregate stability. Methods Soil Anal. Part 1 Phys. Mineral. Prop. Incl. Stat. Meas. Sampl. 9, 511–519 (1965).
Google Scholar 
Rose, C. W. & Rose, C. W. An Introduction to the Environmental Physics of Soil, Water and Watersheds (Cambridge University Press, 2004).
Google Scholar 
Horn, R., Taubner, H., Wuttke, M. & Baumgartl, T. Soil physical properties related to soil structure. Soil Tillage Res. 30, 187–216 (1994).
Google Scholar 
Beven, K. & Germann, P. Macropores and water flow in soils. Water Resour. Res. 18, 1311–1325 (1982).ADS 

Google Scholar 
De Vries, F. T. et al. Abiotic drivers and plant traits explain landscape-scale patterns in soil microbial communities. Ecol. Lett. 15, 1230–1239 (2012).PubMed 

Google Scholar 
Kaisermann, A., de Vries, F. T., Griffiths, R. I. & Bardgett, R. D. Legacy effects of drought on plant–soil feedbacks and plant–plant interactions. New Phytol. 215, 1413–1424 (2017).CAS 
PubMed 

Google Scholar 
Martorell, C., MartÍnez-Blancas, A. & GarcíaMeza, D. Plant–soil feedbacks depend on drought stress, functional group, and evolutionary relatedness in a semiarid grassland. Ecology 102, e03499 (2021).PubMed 

Google Scholar 
Rillig, M. C., Lehmann, A., de Souza Machado, A. A. & Yang, G. Microplastic effects on plants. New Phytol. 223, 1066–1070 (2019).PubMed 

Google Scholar 
Lozano, Y. M. & Rillig, M. C. Effects of microplastic fibers and drought on plant communities. Environ. Sci. Technol. 54, 6166–6173 (2020).ADS 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Denef, K. et al. Influence of dry–wet cycles on the interrelationship between aggregate, particulate organic matter, and microbial community dynamics. Soil Biol. Biochem. 33, 1599–1611 (2001).CAS 

Google Scholar 
Ochoa-Hueso, R. et al. Drought consistently alters the composition of soil fungal and bacterial communities in grasslands from two continents. Glob. Change Biol. 24, 2818–2827 (2018).ADS 

Google Scholar 
Naylor, D. & Coleman-Derr, D. Drought stress and root-associated bacterial communities. Front. Plant Sci. 8, 2223 (2018).PubMed 
PubMed Central 

Google Scholar 
Fu, W. et al. Community response of arbuscular mycorrhizal fungi to extreme drought in a cold-temperate grassland. New Phytol. 234, 2003–2017 (2022).PubMed 

Google Scholar 
Lin, D. et al. Microplastics negatively affect soil fauna but stimulate microbial activity: Insights from a field-based microplastic addition experiment. Proc. R. Soc. B 287, 20201268 (2020).CAS 
PubMed 
PubMed Central 

Google Scholar 
Boots, B., Russell, C. W. & Green, D. S. Effects of microplastics in soil ecosystems: Above and below ground. Environ. Sci. Technol. 53, 11496–11506 (2019).ADS 
CAS 
PubMed 

Google Scholar 
Bosker, T., Bouwman, L. J., Brun, N. R., Behrens, P. & Vijver, M. G. Microplastics accumulate on pores in seed capsule and delay germination and root growth of the terrestrial vascular plant Lepidium sativum. Chemosphere 226, 774–781 (2019).ADS 
CAS 
PubMed 

Google Scholar 
Zimmerman, R. & Kardos, L. Effect of bulk density on root growth. Soil Sci. 91, 280–288 (1961).ADS 

Google Scholar 
Ruser, R., Sehy, U., Weber, A., Gutser, R. & Munch, J. Main driving variables and effect of soil management on climate or ecosystem-relevant trace gas fluxes from fields of the FAM. In Perspectives for agroecosystem
management Chp 2.2, 79–120. ISBN 9780444519054 Elsevier, (2008).Huang, Y. et al. LDPE microplastic films alter microbial community composition and enzymatic activities in soil. Environ. Pollut. 254, 112983 (2019).CAS 
PubMed 

Google Scholar 
Fierer, N., Bradford, M. A. & Jackson, R. B. Toward an ecological classification of soil bacteria. Ecology 88, 1354–1364 (2007).PubMed 

Google Scholar 
Hortal, S. et al. Soil microbial community under a nurse-plant species changes in composition, biomass and activity as the nurse grows. Soil Biol. Biochem. 64, 139–146 (2013).CAS 

Google Scholar 
Scheurer, M. & Bigalke, M. Microplastics in Swiss floodplain soils. Environ. Sci. Technol. 52, 3591–3598 (2018).ADS 
CAS 
PubMed 

Google Scholar 
Jambeck, J. R. et al. Plastic waste inputs from land into the ocean. Science 347, 768–771 (2015).ADS 
CAS 
PubMed 

Google Scholar 
Fuller, S. & Gautam, A. A procedure for measuring microplastics using pressurized fluid extraction. Environ. Sci. Technol. 50, 5774–5780 (2016).ADS 
CAS 
PubMed 

Google Scholar 
Schloerke, B. et al. GGally: Extension to ‘ggplot2’. R package version 2.1. 2. (2021).Venables, W. & Ripley, B. Modern applied statistics with S fourth edition. Publisher Springer-Verlag, New York. (2002).Lenth, R. V. Emmeans: Estimated marginal means, aka least-squares means. R package version 1.6.3 (2021).R Core Team et al. R: A language and environment for statistical computing. (2013).Wickham, H. et al. Welcome to the Tidyverse. J. Open Source Softw. 4, 1686 (2019).ADS 

Google Scholar 
Pedersen, T. L. Patchwork: The composer of plots. R package version 1, 182 (2020).Neuwirth, E. RColorBrewer: ColorBrewer palettes. R package version 1.1-2. (2014). More

Fuhrman JA, Campbell L. Marine ecology – microbial microdiversity. Nature. 1998;393:410–1.CAS 

Google Scholar 
Thompson JR, Pacocha S, Pharino C, Klepac-Ceraj V, Hunt DE, Benoit J, et al. Genotypic diversity within a natural coastal bacterioplankton population. Science. 2005;307:1311–3.CAS 
PubMed 

Google Scholar 
Oh S, Buddenborg S, Yoder-Himes DR, Tiedje JM, Konstantinidis KT. Genomic diversity of Escherichia isolates from diverse habitats. PLoS ONE. 2012;7:e47005.CAS 
PubMed 
PubMed Central 

Google Scholar 
Fuhrman JA, Steele JA, Hewson I, Schwalbach MS, Brown MV, Green JL, et al. A latitudinal diversity gradient in planktonic marine bacteria. Proc Natl Acad Sci USA. 2008;105:7774–8.CAS 
PubMed 
PubMed Central 

Google Scholar 
Sunagawa S, Coelho LP, Chaffron S, Kultima JR, Labadie K, Salazar G, et al. Structure and function of the global ocean microbiome. Science. 2015;348:1261359.PubMed 

Google Scholar 
Raes EJ, Bodrossy L, van de Kamp J, Bissett A, Ostrowski M, Brown MV, et al. Oceanographic boundaries constrain microbial diversity gradients in the South Pacific Ocean. Proc Natl Acad Sci USA. 2018;115:EB266–EB75.
Google Scholar 
Brown MV, Fuhrman JA. Marine bacterial microdiversity as revealed by internal transcribed spacer analysis. Aquat Microb Ecol. 2005;41:15–23.
Google Scholar 
Johnson ZI, Zinser ER, Coe A, McNulty NP, Woodward EMS, Chisholm SW. Niche partitioning among Prochlorococcus ecotypes along ocean-scale environmental gradients. Science. 2006;311:1737–40.CAS 
PubMed 

Google Scholar 
Larkin AA, Blinebry SK, Howes C, Chandler J, Zinser ER, Johnson ZI. Niche partitioning and biogeography of high light adapted Prochlorococcus across taxonomic ranks in the North Pacific. ISME J. 2016;10:1555–67.PubMed 
PubMed Central 

Google Scholar 
Zinser ER, Johnson ZI, Coe A, Karaca E, Veneziano D, Chisholm SW. Influence of light and temperature on Prochlorococcus ecotype distributions in the Atlantic Ocean. Limnol Oceanogr. 2007;52:2205–20.
Google Scholar 
Carlson M, Ribalet F, Maidanik I, Durham BP, Hulata Y, Ferrón S, et al. Viruses affect picocyanobacterial abundance and biogeography in the North Pacific Ocean. Nat Microbiol. 2022;7:570–80.CAS 
PubMed 
PubMed Central 

Google Scholar 
Follett CL, Dutkiewicz S, Ribalet F, Zakem E, Caron D, Armbrust EV, et al. Trophic interactions with heterotrophic bacteria limit the range of Prochlorococcus. Proc Natl Acad Sci USA. 2022;119:e2110993118.CAS 
PubMed 
PubMed Central 

Google Scholar 
Moore LR, Rocap G, Chisholm SW. Physiology and molecular phylogeny of coexisting Prochlorococcus ecotypes. Nature. 1998;393:464–7.CAS 
PubMed 

Google Scholar 
Zwirglmaier K, Jardillier L, Ostrowski M, Mazard S, Garczarek L, Vaulot D, et al. Global phylogeography of marine Synechococcus and Prochlorococcus reveals a distinct partitioning of lineages among oceanic biomes. Environ Microbiol. 2008;10:147–61.PubMed 

Google Scholar 
Carlson CA, Morris R, Parsons R, Treusch AH, Giovannoni SJ, Vergin K. Seasonal dynamics of SAR11 populations in the euphotic and mesopelagic zones of the northwestern Sargasso Sea. ISME J. 2009;3:283–95.CAS 
PubMed 

Google Scholar 
Ribalet F, Swalwell J, Clayton S, Jimenez V, Sudek S, Lin YJ, et al. Light-driven synchrony of Prochlorococcus growth and mortality in the subtropical Pacific gyre. Proc Natl Acad Sci USA. 2015;112:8008–12.CAS 
PubMed 
PubMed Central 

Google Scholar 
Pernthaler A, Pernthaler J, Schattenhofer M, Amann R. Identification of DNA-synthesizing bacterial cells in coastal North Sea plankton. Appl Environ Microbiol. 2002;68:5728–36.CAS 
PubMed 
PubMed Central 

Google Scholar 
Del Giorgio PA, Gasol JM. Physiological structure and single-cell activity in marine bacterioplankton. In: Kirchman DL, editor. Microbial ecology of the oceans. 2nd ed. New Jersey:John Wiley & Sons, Inc.; 2008. p. 243–298.Lin Y, Gazsi K, Lance VP, Larkin A, Chandler J, Zinser ER, et al. In situ activity of a dominant Prochlorococcus ecotype (eHL-II) from rRNA content and cell size. Environ Microbiol. 2013;15:2736–47.CAS 
PubMed 

Google Scholar 
Kirchman DL. Growth rates of microbes in the oceans. Annu Rev Mar Sci. 2016;8:285–309.
Google Scholar 
Landry MR, Selph KE, Yang EJ. Decoupled phytoplankton growth and microzooplankton grazing in the deep euphotic zone of the eastern equatorial Pacific. Mar Ecol Prog Ser. 2011;421:13–24.
Google Scholar 
Selph KE, Landry MR, Taylor AG, Yang E-J, Measures CI, Yang J, et al. Spatially-resolved taxon-specific phytoplankton production and grazing dynamics in relation to iron distributions in the Equatorial Pacific between 110 and 140 degrees W. Deep Sea Res Part II Top Stud Oceanogr. 2011;58:358–77.CAS 

Google Scholar 
Hunt DE, Lin Y, Church MJ, Karl DM, Tringe SG, Izzo LK, et al. Relationship between abundance and specific activity of bacterioplankton in open ocean surface waters. Appl Environ Microbiol. 2013;79:177–84.CAS 
PubMed 
PubMed Central 

Google Scholar 
Sintes E, Herndl GJ. Quantifying substrate uptake by individual cells of marine bacterioplankton by catalyzed reporter deposition fluorescence in situ hybridization combined with microautoradiography. Appl Environ Microbiol. 2006;72:7022–8.CAS 
PubMed 
PubMed Central 

Google Scholar 
Hall E, Maixner F, Franklin O, Daims H, Richter A, Battin T. Linking microbial and ecosystem ecology using ecological stoichiometry: a synthesis of conceptual and empirical approaches. Ecosystems. 2011;14:261–73.
Google Scholar 
Musat N, Foster R, Vagner T, Adam B, Kuypers MM. Detecting metabolic activities in single cells, with emphasis on nanoSIMS. FEMS Microbiol Rev. 2012;36:486–511.CAS 
PubMed 

Google Scholar 
Hatzenpichler R, Scheller S, Tavormina PL, Babin BM, Tirrell DA, Orphan VJ. In situ visualization of newly synthesized proteins in environmental microbes using amino acid tagging and click chemistry. Environ Microbiol. 2014;16:2568–90.CAS 
PubMed 
PubMed Central 

Google Scholar 
Sebastián M, Gasol JM. Visualization is crucial for understanding microbial processes in the ocean. Philos Trans R Soc Lond B. 2019;374:20190083.
Google Scholar 
Korem T, Zeevi D, Suez J, Weinberger A, Avnit-Sagi T, Pompan-Lotan M, et al. Growth dynamics of gut microbiota in health and disease inferred from single metagenomic samples. Science. 2015;349:1101–6.CAS 
PubMed 
PubMed Central 

Google Scholar 
Brown CT, Olm MR, Thomas BC, Banfield JF. Measurement of bacterial replication rates in microbial communities. Nat Biotech. 2016;34:1256–63.CAS 

Google Scholar 
Gao Y, Li H. Quantifying and comparing bacterial growth dynamics in multiple metagenomic samples. Nat Methods. 2018;15:1041–4.CAS 
PubMed 
PubMed Central 

Google Scholar 
Emiola A, Zhou W, Oh J. Metagenomic growth rate inferences of strains in situ. Sci Adv. 2020;6:eaaz2299.CAS 
PubMed 
PubMed Central 

Google Scholar 
Long ANM, Hou SW, Ignacio-Espinoza JC, Fuhrman JA. Benchmarking microbial growth rate predictions from metagenomes. ISME J. 2020;15:183–195.PubMed 
PubMed Central 

Google Scholar 
Carroll J, Van Oostende N, Ward BB. Evaluation of genomic sequence-based growth rate methods for synchronized Synechococcus cultures. Appl Environ Microbiol. 2021;01743-21.Vaulot D, Marie D. Diel variability of photosynthetic picoplankton in the equatorial Pacific. J Geophys Res Oceans. 1999;104:3297–310.CAS 

Google Scholar 
Hunter-Cevera KR, Neubert MG, Olson RJ, Shalapyonok A, Solow AR, Sosik HM. Seasons of Syn. Limnol Oceanogr. 2020;65:1085–102.PubMed 

Google Scholar 
Baer SE, Rauschenberg S, Garcia CA, Garcia NS, Martiny AC, Twining BS, et al. Carbon and nitrogen productivity during spring in the oligotrophic Indian Ocean along the GO-SHIP IO9N transect. Deep Sea Res Part II Top Stud Oceanogr. 2019;161:81–91.CAS 

Google Scholar 
Larkin AA, Garcia CA, Ingoglia KA, Garcia NS, Baer SE, Twining BS, et al. Subtle biogeochemical regimes in the Indian Ocean revealed by spatial and diel frequency of Prochlorococcus haplotypes. Limnol Oceanogr. 2020;65:S220–S32.CAS 

Google Scholar 
Larkin AA, Garcia CA, Garcia N, Brock ML, Lee JA, Ustick LJ, et al. High spatial resolution global ocean metagenomes from Bio-GO-SHIP repeat hydrography transects. Sci Data. 2021;8:1–6.
Google Scholar 
Baym M, Kryazhimskiy S, Lieberman TD, Chung H, Desai MM, Kishony R. Inexpensive multiplexed library preparation for megabase-sized genomes. PLoS ONE. 2015;10:e0128036.PubMed 
PubMed Central 

Google Scholar 
Wandro S, Oliver A, Gallagher T, Weihe C, England W, Martiny JBH, et al. Predictable molecular adaptation of coevolving Enterococcus faecium and lytic phage EfV12-phi1. Front Microbiol. 2019;9:3192.PubMed 
PubMed Central 

Google Scholar 
Oliver A, LaMere B, Weihe C, Wandro S, Lindsay KL, Wadhwa PD, et al. Cervicovaginal microbiome composition is associated with metabolic profiles in healthy pregnancy. mBio. 2020;11:e01851–20.CAS 
PubMed 
PubMed Central 

Google Scholar 
Eren AM, Esen OC, Quince C, Vineis JH, Morrison HG, Sogin ML, et al. Anvi’o: an advanced analysis and visualization platform for ‘omics data. PeerJ. 2015;3:e1319.PubMed 
PubMed Central 

Google Scholar 
Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. Basic local alignment search tool. J Mol Biol. 1990;215:403–10.CAS 
PubMed 

Google Scholar 
van Dongen S, Abreu-Goodger C. Using MCL to extract clusters from networks. In: van Helden J, Toussaint A, Thieffry D, editors. Bacterial molecular networks: methods and protocol. New Jersey:Humana Totowa; 2012. p. 281–295.Bergeron A, Mixtacki J, Stoye J. A unifying view of genome rearrangements. In: Proceeding of the International Workshop on Algorithms in Bioinformatics. Berlin, Heidelberg: Springer; 2006. p. 163–73.Hilker R, Sickinger C, Pedersen CNS, Stoye J. UniMoG-a unifying framework for genomic distance calculation and sorting based on DCJ. Bioinformatics. 2012;28:2509–11.CAS 
PubMed 
PubMed Central 

Google Scholar 
Gelman A, Carlin JB, Stern HS, Dunson D, Vehtari A, Rubin DB. Bayesian data analysis: 3rd ed. Boca Raton:Chapman and Hall/CRC; 2013.Martiny AC, Ustick L, Garcia CA, Lomas MW. Genomic adaptation of marine phytoplankton populations regulates phosphate uptake. Limnol Oceanogr. 2020;65:S340–S50.CAS 

Google Scholar 
Garcia CA, Hagstrom GI, Larkin AA, Ustick LJ, Levin SA, Lomas MW, et al. Linking regional shifts in microbial genome adaptation with surface ocean biogeochemistry. Philos Trans R Soc Lond, B. 2020;375:20190254.CAS 

Google Scholar 
Ustick LJ, Larkin AA, Garcia CA, Garcia NS, Brock ML, Lee JA, et al. Metagenomic analysis reveals global-scale patterns of ocean nutrient limitation. Science. 2021;372:287–291.CAS 
PubMed 

Google Scholar 
Delmont TO, Eren AM. Linking pangenomes and metagenomes: the Prochlorococcus metapangenome. PeerJ 2018;6:e4320.PubMed 
PubMed Central 

Google Scholar 
Worden AZ, Binder BJ. Application of dilution experiments for measuring growth and mortality rates among Prochlorococcus and Synechococcus populations in oligotrophic environments. Aquat Microb Ecol. 2003;30:159–74.
Google Scholar 
Casey JR, Boiteau RM, Engqvist MKM, Finkel ZV, Li G, Liefer J, et al. Basin-scale biogeography of marine phytoplankton reflects cellular-scale optimization of metabolism and physiology. Sci Adv. 2022;8:eabl4930.CAS 
PubMed 
PubMed Central 

Google Scholar 
Berube PM, Rasmussen A, Braakman R, Stepanauskas R, Chisholm SW. Emergence of trait variability through the lens of nitrogen assimilation in Prochlorococcus. Elife. 2019;8:e41043.PubMed 
PubMed Central 

Google Scholar 
Luo HW, Benner R, Long RA, Hu JJ. Subcellular localization of marine bacterial alkaline phosphatases. Proc Natl Acad Sci USA. 2009;106:21219–23.CAS 
PubMed 
PubMed Central 

Google Scholar 
Schapiro JM, Libby SJ, Fang FC. Inhibition of bacterial DNA replication by zinc mobilization during nitrosative stress. Proc Natl Acad Sci USA. 2003;100:8496–501.CAS 
PubMed 
PubMed Central 

Google Scholar 
Hantke K. Bacterial zinc uptake and regulators. Curr Opin Microbiol. 2005;8:196–202.CAS 
PubMed 

Google Scholar 
Glass JB, Axler RP, Chandra S, Goldman CR. Molybdenum limitation of microbial nitrogen assimilation in aquatic ecosystems and pure cultures. Front Microbiol. 2012;3:331.CAS 
PubMed 
PubMed Central 

Google Scholar 
Rusch DB, Halpern AL, Sutton G, Heidelberg KB, Williamson S, Yooseph S, et al. The Sorcerer II Global Ocean Sampling expedition: northwest Atlantic through eastern tropical Pacific. PLoS Biol. 2007;5:e77.PubMed 
PubMed Central 

Google Scholar 
Noell SE, Barrell GE, Suffridge C, Morré J, Gable KP, Graff JR, et al. SAR11 cells rely on enzyme multifunctionality to transport and metabolize a range of polyamine compounds. 2021. https://www.biorxiv.org/content/10.1101/2021.05.13.444117v1.Binder BJ, Chisholm SW. Cell-cycle regulation in marine Synechcococcus sp strains. Appl Environ Microbiol. 1995;61:708–17.CAS 
PubMed 
PubMed Central 

Google Scholar 
Zinser ER, Lindell D, Johnson ZI, Futschik ME, Steglich C, Coleman ML, et al. Choreography of the transcriptome, photophysiology, and cell cycle of a minimal photoautotroph, Prochlorococcus. PLoS ONE. 2009;4:e5135.PubMed 
PubMed Central 

Google Scholar 
Hynes AM, Rhodes KL, Binder BJ. Assessing cell cycle-based methods of measuring Prochlorococcus division rates using an individual-based model. Limnol Oceanogr-Meth. 2015;13:640–50.
Google Scholar 
Vieira-Silva S, Rocha EP. The systemic imprint of growth and its uses in ecological (meta) genomics. PLoS Genet. 2010;6:e1000808.PubMed 
PubMed Central 

Google Scholar 
Weissman JL, Hou SW, Fuhrman JA. Estimating maximal microbial growth rates from cultures, metagenomes, and single cells via codon usage patterns. Proc Natl Acad Sci USA. 2021;118:e2016810118.CAS 
PubMed 
PubMed Central 

Google Scholar 
Chase AB, Karaoz U, Brodie EL, Gomez-Lunar Z, Martiny AC, Martiny JB. Microdiversity of an abundant terrestrial bacterium encompasses extensive variation in ecologically relevant traits. mBio. 2017;8:e01809–17.CAS 
PubMed 
PubMed Central 

Google Scholar 
Olm MR, Crits-Christoph A, Bouma-Gregson K, Firek BA, Morowitz MJ, Banfield JF. inStrain profiles population microdiversity from metagenomic data and sensitively detects shared microbial strains. Nat Biotechnol. 2021;39:727–36.CAS 
PubMed 
PubMed Central 

Google Scholar 
Flombaum P, Gallegos JL, Gordillo RA, Rincon J, Zabala LL, Jiao N, et al. Present and future global distributions of the marine Cyanobacteria Prochlorococcus and Synechococcus. Proc Natl Acad Sci USA. 2013;110:9824–9.CAS 
PubMed 
PubMed Central 

Google Scholar 
Brown MV, Lauro FM, DeMaere MZ, Muir L, Wilkins D, Thomas T, et al. Global biogeography of SAR11 marine bacteria. Mol Sys Biol. 2012;8:595.
Google Scholar  More

The current study is the first to directly compare the differences in eukaryotic community diversity by metabarcoding eRNA and eDNA from an estuarine benthic ecosystem. This is also the first study to compare the diversity of environmental RNA and DNA using two of the most common loci examined in metabarcoding applications: COI and 18S. Only a handful of studies have used eRNA to assess diversity through metabarcoding and have demonstrated some differences in detected diversity between eRNA and eDNA in metabarcoding19,29,38. Environmental DNA has many useful applications, but one of the downfalls of DNA is that it can be detected long after an organism has inhabited an area, such as from dead organisms and even significant distances away from the source1. While the persistence of eDNA may be useful in detecting the presence of endangered, rare, or invasive species in marine systems, the persistence of eDNA, such as legacy DNA, may skew the accurate representation of community structure immediately after a disturbance event or during the exact moment of sampling21. Environmental RNA proves to be a useful tool because it is far less persistent in marine environments2,27. In the current study, RNA provides a snapshot of living organisms present in the mesocosms at the time of sampling, compared to DNA which detects past and present organisms in the sample.Benthic ecosystems often have high biodiversity because they are dynamic environments with various substrates favorable for hosting communities39. To date, there is no one loci and associated primer pair that will effectively detect all eukaryotic organisms, and in particular metazoans, so it is beneficial to use multiple loci when looking for a variety of taxa6. Previous marine eDNA studies have demonstrated the preferred method of using two loci to achieve comprehensive metabarcoding for taxonomically diverse environments5,30,40. The current study further demonstrates the need to use at least two different loci for targeted PCR and sequencing to truly capture the wide diversity in rich systems, such as the top oxygenated layer of the marine benthos. Our results demonstrate the types of organisms detected using the 18S loci vary widely from the organisms detected by COI. The 18S loci also detected a higher number of organisms than using the COI loci. The 18S allowed greater detection of metazoans whereas COI was useful in the detection of Oomycota (i.e., a eukaryotic microorganism that resembles fungi) protozoa. Protozoa are often food sources for meiofauna41; therefore, using COI for protozoan detection and 18S for metazoan detection showcases multiple trophic levels present in the mesocosms of the current study. Utilizing multiple loci not only broadens taxonomic diversity detection but can also highlight multiple trophic levels for understanding food webs originating in benthic environments15.A study evaluating arctic benthic diversity similarly found that COI detected fewer taxa than 18S using eDNA metabarcoding, and there was only ~ 40% taxa overlap between markers at the class level42. In the current study, the COI marker using both nucleic acid templates yielded a higher percentage of unassigned taxa after filtering for presence in the majority of mesocosms compared to 18S. A possible explanation for the low metazoan detection by COI may be that the unassigned taxa are metazoans rather than more SAR organisms. Additionally, singly detected metazoans were filtered out of the analysis if they were not detected in at least 4 mesocosms. For example, one type of amphipod was detected in only 3 mesocosm by RNA using the COI marker, but was not detected in any mesocosms using DNA. Therefore, the amphipod observation was excluded from the COI results in Fig. 3 and Table S3 because it was not found in at least 4 mesocosms.Overall, RNA provides a broader assessment of benthic community structure than DNA, particularly when using two loci/ markers for sequencing. In the current study, we used nuclear 18S ribosomal RNA and DNA and mitochondrial COI RNA and DNA sequences as markers for metabarcoding. The number of copies of ribosomal RNA per cell is higher than the copies of ribosomal DNA, and the ratio of RNA: DNA is higher in single-cell organisms, such as protists43. The higher number of unique ASVs detected using eRNA is likely attributed to the higher number of RNA copies of each marker in small, single-cell organisms successfully amplified during PCR, therefore making rare organisms easier to detect. The DNA of highly abundant or higher biomass organisms may “drown out” (i.e., mask) the sequences from lower abundance and biomass organisms during PCR amplification, thus resulting in lower detected α-diversity. The increased detection of ciliates and protozoa using eRNA are consistent with other recent eRNA metabarcoding results that found higher ciliate and protozoan diversity compared to eDNA using the same Uni18S primers19. Positive correlations between organism biomass and sequence copy numbers have been demonstrated for DNA metabarcoding conducted on invertebrate species44. In the current study, RNA allowed for the detected of both larger meiofauna and smaller microfauna, which is optimal for assessing true biodiversity with molecular assays. Chaetonotida, a type of gastrotrich, was the only taxa detected in 4 of the mesocosms using DNA that was not found with RNA. It is possible the chaetonotida may have died during the experiment due to sensitivities to new environmental conditions, and therefore were not detected with RNA. The temperature of the flowing seawater in the mesocosms was approximately 18 °C, and many marine chaetonotida prefer 23–28 °C and high organic matter substrate45.Previous studies have compared the accuracy of conventional morphological identification to molecular metabarcoding methods for assessing biodiversity. In estuarine-specific studies, metabarcoding methods are able to detect the majority of taxa identified with traditional methods and often detected higher species richness, or higher numbers of unique organisms, that was not found conventionally10,46,47. The limiting factor for higher resolution of taxonomic identification is the availability of species-specific sequences in barcoding databases48. However, barcoding databases are becoming more robust as an increasing number of researchers contribute high quality sequencing data to databases7. Therefore, eRNA metabarcoding techniques perform similarly to conventional morphological methods, and may even uncover higher biodiversity in systems like estuaries where meiofauna have been historically understudied and identified.Although eRNA α-diversity is higher compared to eDNA, there is some overlap between ASVs detected with eRNA and eDNA. The higher percentage of overlap between eDNA and eRNA ASVs is predominantly seen in the ASVs detected in all 7 of the mesocosms. The increased overlap in ASVs detected in all 7 mesocosms compared to the relatively lower overlap of ASVs found in only 1 of the 7 mesocosms is due to the filtering of random organisms found in only 1 mesocosm. Detection of a unique organism in a single mesocosm is likely not representative of the sample community and filters potential artifacts that may be introduced during cDNA synthesis from eRNA. However, all uniquely identified ASVs are used in the β-diversity analysis, so the higher number of unique ASVs detected from eRNA are likely driving the significant differences observed between eRNA and eDNA β-diversity. It is possible that using eRNA could increase the statistical power of a study design compared to eDNA because eRNA detects a higher number of ASVs. It is unlikely that a higher number of RNA ASVs could be due to splice variants contributing to unique sequences because the amplified regions of both markers do not contain introns. Therefore, no splicing of transcripts would be expected. Thus, the detected DNA ASVs are from living organisms in the mesocosms and the higher diversity of ASVs detected from RNA demonstrate that RNA is a more suitable option for assessing diversity of living organisms.It is evident that the mesocosms in the current study were rich with meio- and microfauna due to the number of unique organisms and broad diversity of different taxa. It is likely that collecting samples for nucleic acid extraction directly from the field site may result in higher diversity because there is no mechanical disturbance during the laboratory acclimation period and the presence of other organisms in the system, such as fish, macroinvertebrates, or birds. eDNA molecular abundance in samples has been shown to correlate to actual organismal abundance in laboratory environments, but the same correlation is not as apparent in field samples, which is likely due to a variety of collection and processing methods49. eRNA may be the better nucleic acid template for field collection once flash frozen, especially for the detection of protozoans. Future studies will compare the diversity of eukaryotic communities detected using eDNA and eRNA collected directly from the field rather than from sediment core mesocosms. Repeated sampling from a field site may help reduce transient eRNA detection when establishing accurate baseline community composition in field-based biomonitoring studies.Recently, meiofaunal organisms and communities are being explored as bioindicators, which are organisms whose presence are indicators for environmental stress and pollution17. For example, lower meiofaunal diversity and abundance is associated with higher pollution in harbors, and the presence of some genera of nematodes are correlated with higher concentrations of polycyclic aromatic hydrocarbons because they are more tolerant to pollution50. A previous study that used field-collected mesocosms from the same location as our current study found similar phyla detected with a metabarcoding approach; the majority of the sequences detected in benthic communities were from nematodes, arthropods, and the microfaunal SAR clade10. Similarly, the majority of the ASVs detected from the current study also corresponded with nematodes, arthropods, and the SAR clade, as well as other commonly detected meiobenthic organisms, such as polychaetes and Homalorhagida (i.e., mud dragons). A recent study exposed a benthic foraminiferal community to chromium, and found that eRNA metabarcoding was more robust for detecting changes in diversity at lower chromium concentrations compared to eDNA51. The eRNA metabarcoding method used in the current study detected meiofaunal taxa typical of marine or estuarine environments. Therefore, eRNA metabarcoding may be useful for efficiently identifying bioindicator species or taxa impacted by exposures to different contaminants and environmental stressors to aid with management of aquatic systems. Another advantage of eRNA is that RNA provides functional information about how organisms response to stress through altered transcription of activated pathways. eRNA will likely be a more powerful tool than eDNA because it allows for the detection of both bioindicator species and, in the future with increase development of genetic databases, environmental detection of biomarkers of stress through increased transcription of response genes.Many academic researchers are adopting molecular methods using High Throughput Sequencing as the future of biomonitoring surveys; however, few regulatory and environmental management organizations/ agencies have adopted metabarcoding into their routine biomonitoring practices for regulatory purposes. In marine benthic communities, metabarcoding provides a comprehensive assessment of diversity and is useful for detecting a broader array of organisms in biomonitoring surveys especially when using two markers47. eDNA is also useful for monitoring discrete communities (i.e., benthic versus pelagic)52, so it is possible that using eRNA could provide vital information about living organisms in specific environments compared to eDNA. Metabarcoding sequencing and bioinformatic approaches for benthic environments vary among studies, thus requiring some standardization between methods to further advance the use of metabarcoding in conservation and regulatory applications30,53.Like eDNA, some advancements must be made with eRNA to be used as a quantitative tool in molecular ecology. Validation of eDNA metabarcoding for assessing relative abundance of species is rapidly progressing by correlating laboratory studies of DNA shedding with field experiments1,54. Similar validation techniques can be used to develop eRNA metabarcoding as a quantitative or semi-quantitative method. There is growing interest in optimizing eRNA extraction protocols from different types of environmental media to standardize the use of eRNA in downstream molecular applications55. Thus, standardizing eRNA protocols will help with integration into environmental management toolkits for regulatory purposes.RNA poses unique barcoding challenges compared to DNA because the number of RNA transcripts from a gene are not always present in the same proportion compared to the gene copy number per genome (i.e., one DNA copy per cell), especially for differentially expressed genes. However, one possible way to work around this issue is to utilize constitutively expressed marker loci where transcription is generally stable and unaffected by environmental stressors, such as those used as reference genes for quantitative real-time PCR49. Fortunately, many loci chosen for metabarcoding purposes fit this criterium; the transcription of 18S and COI remain steady within the cell regardless of environmental stress.Collecting sediment cores from the environment and bringing them into controlled laboratory settings for community analysis through eRNA metabarcoding is a powerful tool that opens opportunities for this method to be used in a broad range of fields. For instance, field-collected mesocosms could be used in controlled settings to investigate the effects of individual or mixtures of toxicants on entire community and population-level outcomes. Marine sediments are often the ultimate sink for environmental contaminants, such as organic pollutants , heavy metals56, and plastic particles57, yet few studies investigate actual community-level changes in contaminant exposures. Marine benthic environments have high biodiversity, but the breadth of diversity in micro- and meiofaunal organisms is often understudied because traditional morphological methods are immensely time consuming. Sediment core mesocosms could also be used to understand the effects of global climate change stressors, such as fluctuating temperatures, surface water salinity and pH, on communities as well as conventional and emerging contaminants in combination with climate change stressors. Additionally, this method could also be used to understand how communities respond to a significant disturbance event or smaller series of stressors, which are often difficult to measure in environmental settings32,58. In these applications, eRNA is favorable for constructing community composition at a specific moment of the experiment to better regulate anthropogenic causes of environmental stress. More

Electrohydrogenesis effluent as a potential biofertilizerTo characterize the electrohydrogenesis left-over residues as potential biofertilizers, the sample from the operating reactors was performed a 16S rRNA sequencing test, and interestingly, the results revealed that the bio-electrohydrogenesis effluent was enriched with various microorganisms including plant growth-promoting microbes that display biofertilizer-like features. Among the well-known plant-promoting bacterial genera observed in DF-MEC residues included Azospirillum, Mycobacterium, Chryseobacterium, Paenibacillus, Rhizobacter, Pseudomonas, Achromobacter, Bradyrhizobium, Actinomyces, Sphingomonas, Allorhizobium-Neorhizobium-Pararhizobium-Rhizobium, Gordonia, Rhodococcus, Bacillus, Methylobacterium-Methylorubrum, Microbacterium, Flavobacterium, Devosia, Acinetobacter, Mesorhizobium, Enterobacter, Aeromonas, Beijerinckia, etc.24,25,26 (Fig. 2). Lots of investigations working on the feasibility of using biofertilizers other than chemical fertilizers have revealed that those aforementioned microbes play a major role in providing the required nutrients for enhanced crop yield.Figure 2The abundance of the Plant growth-promoting bacteria (genus level) detected from the DF-MEC digestate (%).Full size imageNitrogen-fixing microorganismsThe detected nitrogen-fixing microorganisms from the electrohydrogenesis effluent include Azospirillum sp. (0.11 ± 0.02%), rhizobia (Rhizobium (0.058 ± 0.02%), Bradyrhizobium (0.11 ± 0.04%), and Mesorhizobium (0.1 ± 0.03%)), and Beijerinckia (0.08 ± 0.03%) (Fig. 2) and were repeatedly reported for their superior contribution to the plants’ nitrogen requirements through biological nitrogen fixation, which is an important component of sustainable agriculture25. Although the atmosphere counts about 78% N2, it couldn’t be used by plants in its natural state. Prior to getting used by plants, it needs to be converted to ammonia, which is the readily assimilable form of nitrogen by plants/or crops via a biological nitrogen fixation mechanism25. The biological Nitrogen fixation mechanism is summarized in Fig. 3.Figure 3Mechanism of nitrogen fixation bio-catalyzed by nitrogenase enzyme. The plant growth-promoting bacteria produce nitrogenase which is a complex enzyme consisting of dinitrogenase reductase and dinitrogenase. This complex enzyme plays a major role in molecular N2 fixation. Dinitrogenase reductase provides electrons and dinitrogenase uses those electrons to reduce N2 to NH3. However, oxygen is a potential threat to this process since it has the ability to get bound to the enzyme complex and make it inactive and consequently inhibit the process. Interestingly, bacterial leghemoglobin has a strong affinity for O2 and thus gets bound to free oxygen more strongly and effectively to suppress the available oxygen effects on the whole process of nitrogen fixation.Full size imagePhosphate-solubilizing microorganismsFurthermore, various phosphate-solubilizing and mineralizing strains were also found in bioelectrohydrogenesis residues collected from our DF-MEC integrated reactors. Among those microorganisms with the ability to solubilize/metabolize the insoluble inorganic phosphorus, the dominant bacterial genera included Pseudomonas (0.65 ± 0.15%), Bacillus (0.44 ± 0.11%), Rhodococcus (0.04 ± 0.009%), Rhizobium (0.05 ± 0.02%), Microbacterium sp. (0.04 ± 0.01%), Achromobacter (0.16 ± 0.07%), and Flavobacterium (0.058 ± 0.014%) (Fig. 2). Though enormous amounts of phosphorus are available in the soil, its high portion never contributes to plant growth in its primitive state, unless it is bio-transformed into absorbable forms including monobasic and dibasic. Microbial phosphate solubilizing mechanisms are well described in Fig. 4.Figure 4Inorganic phosphorus solubilization by phosphate-solubilizing rhizobacteria. A bacterium solubilizes inorganic phosphorus through the action of low molecular weight organic acids such as gluconic and citric acids. The hydroxyl (OH) and carboxyl (COOH) groups of these acids chelate the cations bound to phosphate and thus convert insoluble phosphorus into a soluble organic form. The mineralization of soluble phosphorus occurs by synthesizing different phosphatases which catalyze the hydrolysis process. When plants incorporate these solubilized and mineralized phosphorus molecules, eventually, overall plant growth and crop yield significantly increase.Full size imagePhytohormone-producing microorganismsIn this current work, the electrohydrogenesis effluent also contained bacterial genera such as Mycobacterium (0.77 ± 0.18%), Allorhizobium (0.05 ± 0.02%), Pararhizobium (0.05 ± 0.02%), Paenibacillus (1.18 ± 0.24%), Bradyrhizobium (0.11 ± 0.04%), Rhizobium (0.05 ± 0.02%), Acinetobacter (0.14 ± 0.02%), and Azospirillum (0.11 ± 0.025%) (Fig. 2) that have the ability to synthesize indole-3-acetic acid/indole acetic acid (IAA) through indole-3-pyruvic acid and indole-3-acetic aldehyde25. IAA is a well-known type of phytohormone that enhances plant/crop growth. Particularly, Azospirillum sp., also produce various phytohormones namely cytokinins, gibberellins, ethylene, abscisic acid and salicylic acid, auxins, vitamins such as niacin, pantothenic acid, and thiamine. The conceptional model delineating the positive effects of inoculation with Azospirillum sp. a phytohormones-producer plant growth-promoting rhizobacteria and its detailed functions on plant growth are summarized and illustrated in Fig. S1. Therefore, the existence of those rhizobacteria in the bioelectrohydrogenesis residues further implies the suitability of considering the DF-MEC left-over residues as potential biofertilizers.Heavy metals-bioremediating microorganismsSome other bacterial genera with the ability to bioremediate the heavy metal toxicity were also found within the bioelectrohydrogenesis left-over residues as well. Among the detected plant growth-promoting bacterial genera; Rhizobium (0.058 ± 0.023%), Mesorhizobium (0.1 ± 0.026%), Bradyrhizobium (0.11 ± 0.04%), Pseudomonas (0.65 ± 0.15%), and Achromobacter (0.16 ± 0.077%) were reported for their key contribution to alleviate the toxicity of the heavy metals via bioremediation process and improve the soil quality for a relief plant development26 (Fig. 2). Other detected heavy metals-bioremediating microorganisms’ species were Chryseobacterium sp. (0.08 ± 0.007%), Azospirillum (0.11 ± 0.02%), Bacillus (0.44 ± 0.11%), Enterobacter (8.57 ± 0.9%), Gordonia (0.06 ± 0.02%), Paenibacillus (1.18 ± 0.24%), Pseudomonas (0.65 ± 0.15%), and Actinomycetes (0.36 ± 0.05%) that either use microbial siderophores or enzymatic biodegradation process.Electrohydrogenesis left-over residues as a potential source of essential elements for plant growthAs aforementioned in “Materials and methods” section, the electrohydrogenesis left-over residues contained diverse microbial communities that degraded the MEC substrate and generate biogas and inorganic compounds. Moreover, it has been reported that those inorganic nutrients are generally available in fermentation effluent in readily plant-utilizable formats owing to substrate mineralization27. Beside detecting various plant growth-promoting microorganisms in the electrohydrogenesis effluent, a larger number of mineral elements essential for promoted growth and development of crop plants were also investigated and analyzed from the residues. The detected primary and secondary macro-elements’ concentrations in the residues were arranged in decreasing order as follows P  > S  > Na  > K  > N  > Ca  > Mg. Interestingly the findings show that the residues abundantly contained Phosphorus (2.766 × 103 mg/L), Nitrogen (274 mg/L), Potassium (282 mg/L), Calcium (17.66 mg/L), Magnesium (16.3 mg/L), Sulfur (1.225 × 103 mg/L), and Sodium (294.3 mg/L) which are well known as macro-nutrients needed in larger amounts for enhanced plant/ crop growth (Fig. 5).Figure 5Macro-, and micronutrients detected from the bio-electrohydrogenesis left-over residues (mg/L).Full size imageMoreover, small amounts of the microelements including Ni, Pb, Zn, Cu, Cr, Hg, Cd were also found in the electrohydrogenesis residues, and consistently these elements are generally required in small quantities for the development of plants (Fig. 5), otherwise, their high concentrations are toxic for the plant cells thus suppress or inhibit plant growth. The detected concentrations for the main microelements in this current research ranged only from 0.36 to 9.6 × 10–5 mg/L and were all reported to play fundamental roles in plant metabolic reactions.Cultivation of the leguminous crops using electrohydrogenesis left-over residues as fertilizerAfter evaluating the plant-growth promoting bacterial communities and the macro- and micronutrients required for plant/crop growth in the electrohydrogenesis left-over residues, the latter was directly used as fertilizer to grow three different plant species including tomato, chili, and brinjal as afore-described in the “Materials and methods” section. To access the potentials of the electrohydrogenesis effluent as fertilizer, the plants grown in the soil amended with the effluent (Soil + Effluent), were directly compared with their corresponding control plants (Soil + water). The results indicated that at the end of 1st month, the plants with effluent grew faster and generated a good amount of branching than the control plants (see Fig. 6), possibly due to the availability of both microbial species with bio-fertilizing aspects and micro-and macronutrients in the effluent.Figure 6Analysis of the plant growth at the end of the 1st month of cultivation. (a) Tomato in soil with effluent, and its control without effluent (b); (c) Chilli grown in soil with effluent, and its control without effluent (d); and (e) brinjal grown in soil with effluent, and its corresponding control grown without effluent (f) (after 2 months).Full size imageFor instance, tomato (Solanum lycopersicum L.) and chilli (Capsicum annuum L.) height in soil + electrohydrogenesis effluent was ~ 36.9 ± 2.1 cm and ~ 32.6 ± 0.8 cm respectively which was ~ 2.03 and ~ 1.2 times the height of their corresponding plant species in the control protocol, respectively (see Fig. 7). However, the brinjal species (Solanum melongena L.) didn’t show any remarkable height differences in both protocols after a month of cultivation (data not shown), probably due to their low adaptative characteristics to the new environment. However, after the 2nd month, the brinjal height in soil + effluent became 2.7 times that of the brinjal control cultivated without effluent (see Fig. 6e,f). Moreover, both the number of the plants’ leaves and their length in plants cultivated in soil + effluent, were remarkably higher than in plants grown without the supply of the effluent.Figure 7Daily plant growth analysis within one month of cultivation. (a) Tomato growth monitoring, (b) Chili growth analysis.Full size imageAt the end of the 3rd month, the plants in soil + electrohydrogenesis effluent generated more fruit with big size than the control plants (see Fig. 8), but the tomato (Solanum lycopersicum L.) didn’t generate fruits in both protocols at that time probably due to the high weather temperature that inhibitory affected its continuous growth, as previously reported that tomato species are generally so sensitive to temperature change28,29. The final yield was evaluated in terms of the size and number of fruits per cultivated plant. Chili cultivated in soil with MEC effluent generated 3 fruits/plant and its corresponding control without effluent produced only 1 fruit/plant. The chili fruit size in soil + effluent was 16 cm, approximately 18.7% higher than its corresponding control. Moreover, at the time of collecting data, the brinjal plant cultivated in soil with MEC effluent generated brinjal fruits whereas its corresponding cultivated without electrohydrogenesis effluent started flowering (see Fig. 8). These further indicate the significant contribution of the electrohydrogenesis effluent in speeding up the plant growth. Herein, the electrohydrogenesis left-over residues have notably improved the soil quality and significantly promoted the plants’ phenology characterized by plant growth, the generation of new leaves, flowering, and the production of fruits.Figure 8Analysis of plant growth characterized by the flowering and fruiting process at the end of the 3 months. (a) Chili grown in soil with effluent, and its control without effluent (b); (c) brinjal grown in soil with effluent and its corresponding control grown without effluent (d).Full size image More

We applied a modified version of a planning framework for CWR conservation25,26 which has been used by numerous countries of Europee.g.29,63,64, Americae.g.65, Africa30 and Asia66,67. We addressed the following main steps of the toolkit (see Spanish version49): (i) CWR checklist, i.e., creating a list of CWR taxa distributed in an area (Supplementary Data 1), (ii) CWR inventory, i.e., taxa selection and collation of ancillary data, including taxonomic data (Supplementary Data 2), (iii) taxa extinction risk assessment (Table 1, Supplementary Data 3), and (iv) a systematic conservation planning assessment, i.e., spatial analyses to assess conservation areas (Fig. 1). We only provide a brief description of steps i-iii, as these are thoroughly described in Goettsch et al.2. Here, we focus on the systematic conservation planning assessment, introducing an approach in order to identify conservation areas for CWR that account for genetic differentiation in a spatially explicit way, through the use of proxies of genetic differentiation (Fig. 1).During the process -framed under the project “Safeguarding Mesoamerican crop wild relatives” (https://www.darwininitiative.org.uk/project/23007/)- more than 100 experts from academic, governmental, and non-governmental organizations from El Salvador, Guatemala, Honduras, Mexico, the UK, and IUCN participated in six workshops, shared data, and provided fundamental knowledge and feedback at each project stage to ensure accurate, reliable and robust information for next steps. The checklist, inventory and risk assessment were collaboratively developed between partners of El Salvador, Guatemala, and Mexico (hereafter, Mesoamerica; Goettsch et al.2). The spatial analysis to identify areas for in situ and ex situ conservation of CWR was done independently by each country.To assess conservation areas of CWR in Mexico, we developed proxies of genetic differentiation that account for evolutionary processes by including historical and environmental drivers of genetic diversity (see the Methods section ‘Proxies of genetic differentiation’). In addition, we used criteria such as information on taxon-specific tolerance to human-modified habitats and IUCN extinction risk category. We applied a systematic conservation planning approach and performed spatial analysis using the software Zonation50. We compared different scenarios to represent genetic diversity of CWR based on potential species distribution models (SDM) and proxies of genetic differentiation.Study areaMesoamerica is a cultural region encompassing the territories of Belize, Guatemala, El Salvador, the southern part of Mexico and parts of Honduras, Nicaragua and Costa Ricasee 2. In this study, we also included the dry areas of northern Mexico that are part of Aridamerica68 and the Nearctic biogeographic realm69 to account for the full extent of the geographic range of many taxa included in the extinction risk assessment2.For the assessment of conservation areas, we focused on Mexico, which is one of the most biodiverse countries in the world70. The Mexican territory covers 80% of the landscapes of the region called Mesoamerica. Its high biological diversity is attributed to its geographic, topographic, climatic, geological and cultural characteristics, which, among other factors, shaped the distribution of an extraordinary variety of ecosystems and species with high levels of endemism and species turnover among different regions32,71,72,73. In particular, the high genetic variation within populations of landraces and CWR is the result of past and ongoing sociocultural processes occurring in a wide range of distinct environmental conditions74,75.(i) CWR checklist and (ii) CWR inventoryThe compiled CWR checklist included ~3000 species and subspecies of 92 genera and 45 families of plants that belong to the same genus of a crop cultivated in Mesoamerica, or wild plant collected for food or other uses in the region (Supplementary Data 1).The first set of criteria were established in preparation for the first stakeholder workshop. The following criteria were applied at the genus level to compile the CWR inventory: (1) occurrence of wild relatives of cultivated plants or crops that were domesticated in Mesoamerica; (2) existence of research groups working on taxa that could support the extinction risk assessment; and (3) relation to a crop of economic and nutritional importance at local, national and regional levels, or cultivars known to require genetic improvement.To narrow the list for the inventory and extinction risk assessment, similar criteria were agreed upon in the same workshop and applied at the species level: (1) native distribution in Mesoamerica, incl. Aridamerica; (2) related to a crop of economic or social importance based on production and nutritional value; (3) related to a taxon for which Mesoamerica is the center of origin or domestication; (4) constitutes part of the primary or secondary gene pool, and in some cases the tertiary gene pool76. The primary gene pool consists of wild plants of the same species as the crop and thus their mating produces strong fertile progeny. The secondary gene pool is composed of wild relatives distinct from cultivated species but closely related as to produce some fertile offspring (same taxonomic series or section in the absence of crossing and genetic diversity information, see the ‘taxon group’ concept proposed by Maxted and collaborators77, Supplementary Note 5). The tertiary gene pool (same subgenus in the taxon group concept) corresponds to CWR that are more distant relatives to the taxa of the primary gene pool, but can have important adaptive traits which can be used with specific breeding techniques. This provided a preliminary list of 514 CWR taxa related to avocado, cotton, amaranth, cocoa, squash, sweet potato, chayote, chili pepper, cempasuchil, bean, sunflower, maize, papaya, potato, vanilla, and yuca (Supplementary Data 2).The list had to be further reduced due to time and funding restrictions to include those genera which when added together would include no more than 250 taxa, and that the taxonomic groups could be comprehensively assessed and their taxa evaluated throughout their entire range. Thus, not all species in the group necessarily met the criteria previously mentioned. See the final Mesoamerican CWR inventory in Supplementary Data 3; see summary in Table 1.(iii) Taxa extinction risk assessmentFull methodological details and results of this section are described in Goettsch et al.2. Summarizing, during the process 224 taxa were evaluated according to the International Union for Conservation of Nature, IUCN, Red List Categories and Criteria78. The IUCN Red List is a critical indicator to identify species most vulnerable to extinction considering a set of criteria, i.e., species’ population trends, size, structure, and geographic ranges. A Red List workshop with the participation of 25 experts from different project partner institutions and IUCN specialists was organized to assess the extinction risk of taxa. The threat analysis included not only species, but subspecies and subpopulations (i.e. races) for some groups (Supplementary Data 3, see summary in Table 1).(iv) Systematic conservation planning assessmentTo undertake the following spatial analyses we focused on the dataset of 224 CWR described above, which is representative of the CWR of the main crops of Mesoamerica (10 genera, Table 1).Species distribution modelingTo compile occurrence records, hundreds of data sources were consulted, including published and personal databases of the project participantse.g.79,80,81,82, the Agrobiodiversity Atlas of Guatemala (https://www.ars.usda.gov/northeast-area/beltsville-md-barc/beltsville-agricultural-research-center/national-germplasm-resources-laboratory/docs/atlas-of-guatemalan-crop-wild-relatives), the Global Biodiversity Information Facility (GBIF, https://www.gbif.org/), and Mexico’s Biodiversity Information System (SNIB, http://snib.mx/).To generate potential species distribution models (SDM), we used more than 13,000 occurrence records (Supplementary Data 4), that were standardized and curated by experts to generate the range maps of taxa as part of the extinction risk assessment, which were published in IUCN Red List (https://www.iucn.org/news/species/202109/threats-crop-wild-relatives-compromising-food-security-and-livelihoods). Spatial resolution of the SDM was 1 km2. SDM were obtained for taxa with more than 20 unique occurrence data in a 1 km2 grid covering the study extent to reduce uncertainty when using smaller sample sizes83. We used 19 bioclimatic variables and other climatic variables, such as annual potential evapotranspiration, aridity index, annual radiation, slope, and altitude84,85,86. Climate data represents annual and seasonal patterns of climate between 1950 and 2000. Also, we used a variable that described the percentage of bare soil and cultivated areas87. Collinearity between variables was assessed with the ‘corselect’ function of the package fuzzySim version 1.088, using a value of 0.8 and the variance inflation factors as criteria to exclude highly correlated variables.We used MaxEnt version 3.3.1, a machine-learning algorithm that uses the maximum entropy principle to identify a target probability distribution, subject to a set of constraints related to the occurrence records and environmental data89,90. Model calibration area for each taxon included those ecoregions where the taxon has been recorded; we used the terrestrial ecoregions dataset69. We did this based on the calibration area or ‘M element’ of the BAM diagram that refers to areas that have been accessible to the taxon via dispersal over relevant periods of time91,92. We randomly sampled 10,000 background localities from the selected areas.To reduce model complexity without compromising model performance, we built several models by varying the feature classes (FC) and regularization multipliers (RM) (see refs. 93,94,95) using R 3.6.096 and ‘ENMeval’ version 0.3.0 package97. FC determines the flexibility of the modeled response to the predictor variables, while the RM penalizes model complexity93. Occurrence records were randomly divided into 70% for model selection, and 30% of data was withheld for model validation. ENMeval carries out an internal partition of localities to test each combination of settings. Therefore, we selected the random k-fold method to divide localities into four bins. We build models with six FC combinations and varied RM values ranging from 0.5 to 4.0 in 0.5 increments. Optimal models were selected using Akaike’s Information Criterion corrected for small sample sizes (⍙AICc = 0). This method penalizes overly complex models and helps to choose those with an optimal number of parameters. However, it has been shown that the number of model parameters may not correctly estimate degrees of freedom98, and that model selection should not be selected solely with one measure99. Thus, we used 30% of the withheld data to test the area under the curve (AUC) of the receiver operating characteristic, and the omission error under a 10 percentile training threshold.We used the ten percentile or minimum training presence threshold to obtain binary maps of the presence and absence of suitable areas for species distribution. We asked experts of each taxonomic group who were also involved in the extinction risk assessment to select one of these two options and to indicate possible overestimated areas, which were then eliminated case by case using the information of Mexican ecoregions100 and watersheds101. Eight models were binarized with the minimum training presence threshold; for the other models we used the 10 percentile threshold. See MaxEnt performance and significance of SDM at Supplementary Data 5. AUC values ranged from 0 to 1; 0.5 indicated a model performance not better than random, while values closer to 1 indicated a better model performance; here we used SDM showing AUC values higher than 0.7. For Phaseolus and Zea, we used SDM that were previously generated by Delgado-Salinas et al.102, and Sánchez González et al.103, respectively. SDM for 116 taxa were validated by experts of each taxonomic group. See references and download links at Supplementary Data 6.For the conservation planning analysis of Mexico, we clipped the models to the Mexican territory, and trimmed the continuous SDM using the binary SDM to keep pixel values of areas with elevated probability of taxa presence. For taxa without SDM, we included the occurrence records of these taxa in the spatial analysis by using the information on observation location, i.e., coordinates (see Supplementary Data 3). This is done by enabling the function ‘species of special interest’ (SSI). See further details in the method section ‘Final conservation analysis’.Proxies of genetic differentiationTo identify proxies of genetic differentiation in an explicit, efficient, and repeatable way, we included environmental and historical drivers of genetic diversity. For this, we first divided Mexico into 27 Holdridge life zones (Supplementary Fig. 2, Supplementary Data 8), which we then subdivided according to phylogeographic studies that have found genetic differentiation among populations of several taxa (see division of each life zone into proxies in Supplementary Fig. 4; Supplementary Fig. 3 provides a general geographical overview of Mexico and main geographic references mentioned in Supplementary Fig. 4). The literature review was done searching for the words “phylogeography” and one of the following: (i) name of the Mexican biogeographic zones, (ii) “Mexico” + an ecosystem name (e.g. “Mexico” “rainforest”) or (iii) “Mexico” + lowlands/highlands. See list of references used in this study in Supplementary Data 9.In addition, we manually reviewed the citations to the most cited papers of the previous search. Reviews and meta-analyses were also included, although we excluded studies performed in CWR to show that our approach can be used without prior information on this group. As more studies on such taxa become available, they can be used to fine-tune the proxies of genetic differentiation. We focused on terrestrial species including plants, animals, and fungi (Supplementary Data 10) except to subdivide a life zone covering the coasts of the California Peninsula, where we could not find studies on terrestrial taxa so we included studies on fish species (see Supplementary Fig. 4).Since most of the life zones cover large territories, and complete phylogeographic congruence among different taxa is uncommon, we targeted to represent general trends that would likely occur across diverse species, instead of trying to represent fine idiosyncratic patterns of genetic differentiation. For instance, although distribution ranges of highland taxa shifted during the Pleistocene climate fluctuations, in general populations persisted (glacial-interglacial periods) within the main mountain ranges, while lowland populations were ephemeral (only glacial periods). So, gene flow among mountain ranges was more limited than within them. As a result, genetic differentiation among mountain ranges of different biogeographic provinces has been widely documented32, so we used this general pattern to subdivide the life zones that occur in highlands. These types of patterns are particularly relevant for a country like Mexico, due to its complex topography, tropical latitude, and geographic features of different ages, which promote population differentiation among the Mexican main geographic features. To translate the phylogeographic information into a spatial context, we used biogeographic regions, basins, topographic or edaphic data to split the life zones into different subzones using the best fitting cartography to represent the phylogeographic patterns (Supplementary Fig. 4).We obtained 102 proxies of genetic differentiation for Mexico (Supplementary Fig. 5). We validated our findings by using available genomic data of an empirical study of a wild relative of maize, the teosinte Zea mays subsp. parviglumis, which was not included in the literature review in order to test the usefulness of our approach regarding the lack of genetic data. The dataset includes ca. 1800 occurrence records and ca. 30,000 SNPs48. Sampling localities were not used for distribution modeling. Admixture groups per population were estimated for K1 to 60. According to the population analysis, Z. mays subsp. parviglumis is structured in 13 genetic clusters along a longitudinal gradient (Fig. 3a–c). We used the K = 13 for plotting based on the Cross-Validation error. The proportion of each genetic cluster was estimated by sampling locality and plotted using pie charts over the map (Supplementary Fig. 6). Then, using the data layer of the SDM subdivided by proxies of genetic differentiation, we extracted which was the proxy most frequent in a 5 km buffer for each sampling locality. The Admixture plot was ordered by all genetic clusters and subdivided by the proxy of genetic differentiation most frequent for each locality. In addition, we calculated a principal component analysis (PCA) and projected into a score plot the first three components. Individual samples were colored by the proxies where they fell in the 5 km buffer (Fig. 3c). To compare how genetic variation was represented by the different scenarios we plotted the proportion of the area of each proxy as given by the potential SDM according to two different scenarios (only considering SDM; combining SDM*PGD) considering 20% of Mexico’s terrestrial area (Fig. 3d). Analyses were run in R version 3.5.196 using the R packages pcadapt version 4.3.3104, ggplot2 version 2_3.3.3105, readr version 1.4.0106, gridExtra version 2.3107, ggnewscale version 0.4.5108, scatterpie version 0.1.5109, pophelper version 2.3.1110, raster version 3.4-5111, rgdal version 1.4-8112, rgl version 0.107.10113, and sp version 1.4-4114,115.Habitat preferenceWe considered habitat preference to refine the presence of CWR in the planning process; thus minimizing commission errors and highlighting areas that more probably contain taxa116. For each taxon, experts assessed its habitat preference (1: high preference; 0.5: low preference; 0.1: no preference) according to the following categories: (i) well-conserved vegetation (i.e. primary vegetation), (ii) human-impacted vegetation (i.e. secondary vegetation), (iii) less intensive rainfed and moisture agriculture, (iv) intensive rainfed and moisture agriculture, (v) irrigated agriculture, (vi) induced and cultivated grasslands and forests, and vii) urban areas (Supplementary Data 11). To spatially delimit these classes, we used the land use cover and vegetation map for Mexico117, and assessed seven main categories of land cover by grouping the map legend (Supplementary Fig. 9). To differentiate between less intensive and intensive cultivated areas, we followed Bellon et al.56, who associated the presence of native maize varieties of Mexico to occur in municipalities with average yields of less than or equal to 3 t ha-1 using agricultural production data from 2010 from the Information System of Agrifood and Fisheries (SIAP), and selected the municipalities with the established average maize yield. We combined the municipality layer with the land cover map to differentiate areas of high and low agricultural intensity. To generate taxon-specific habitat layers, we associated the habitat preference classes established by experts to the land cover map aggregated into seven major land cover categories, using R 3.6.096 and the following packages: raster version 3.4-5111 and rgdal version 1.4-8112. We obtained habitat maps for 116 taxa with SDM.Preliminary analysisWe generated five preliminary scenarios to explore different approaches to include conservation features for maximizing the representation of intraspecific diversity as given by taxa and proxies of genetic differentiation, i.e., representation of proxies within a taxa range (Supplementary Fig. 7): (i) “SDM” scenario, included 116 SDM, which we used as base scenario to examine the representation of taxa and proxies of genetic variability (n = 116); (ii) “SDM + LZ” scenario, included 116 SDM and 27 layers representing Holdridge life zones to consider environmental variation (n = 143); (iii) “SDM + PGD” scenario, included 116 SDM and 102 layers representing each proxy of genetic differentiation individually (n = 218); (iv) “SDM*PGD” scenario, included 5004 input layers representing the intersection of SDM and PGD (n = 5004; combining 116 SDM with 102 proxies resulted in 11,832 layers, but as some of the intersections produced empty outputs given the extension of SDM that do not cover all Mexico, for further analysis we used 5004 input layers with value data. To subdivide the layers, we used ArcGIS version 10.2.2118; to filter the layers, we used R 3.5.196.); (v) “SDM and PGD as ADMU” scenario, included 116 SDM as the main conservation features, while integrating one single layer of proxies of genetic differentiation to consider each of them as planning units by using the ‘Administrative units’ function. Analysis was done in Zonation50,119.We compared the results by assessing 20% of Mexico’s terrestrial area (Fig. 5b) to perform statistical analysis in R 3.5.196 using the following packages: purrr version 0.3.4120, ‘dplyr’ version 1.0.2121, ‘ggplot2’ version 2_3.3.3105, ‘raster’ version 3.4-5111, ‘scales’ version 1.2.0122, ‘sp’ version 1.4-4114,115, ‘tidyr’ version 1.0.2123, and ‘vegan’ version 2.6-2124. The area threshold was established based on Aichi target 11 and on comparisons of performance curves to efficiently represent taxa ranges delimited by SDM and proxies of genetic differentiation (Fig. 6). As using SDM combined with proxies of genetic differentiation showed the highest representation of genetic diversity (“SDM*PGD” scenario), we used this approach for the final analyses.Final conservation analysisWe identified areas of high conservation value for CWR in Mexico by using the software Zonation version 4.050,119, a systematic conservation planning tool that allows optimizing representation of species, taxa, or other conservation features, e.g., proxies of genetic differentiation, in a given study area. The program hierarchically ranks areas by removing cells of low conservation value, as given, for example, by a reduced number of taxa or occurrence of low weighted features, while considering multiple criteria such as the weighting of taxa and habitat preference of taxa. We applied the core-area zonation removal rule (CAZ) to maximize the representation of all conservation features in a minimal possible area51. Zonation generates two main outputs: (a) a hierarchical landscape priority rank map, that allows decision makers establishing different area thresholds to highlight areas of conservation interest; and (b) a representation curve showing species or conservation features range distribution in a given area. The curve also allows identifying how much area is needed to cover a certain taxon range or the distribution of a feature of conservation interest.For the conservation scenarios, we integrated the following inputs in the Zonation software: (1) 5,004 layers, i.e., SDM intersected with proxies of genetic differentiation (as described by “SDM*PGD” scenario, Fig. 4), (2) occurrence records of 98 taxa; only for those taxa without SDM, see Supplementary Data 3), (3) taxa specific habitat layers (according to Supplementary Data 11 and Supplementary Fig. 9), and (4) IUCN threat category (Supplementary Data 3) as an additional parameter to weight taxa differently to consider their vulnerability to extinction, see details below. See Zonation configuration at Supplementary Note 6.Data from different sources can be mixed in the same analysis, which is useful to not lose or omit information of any taxa of interest in the assessment. Here, we included information of a total of 214 taxa (see Supplementary Data 3). Distribution data of 116 taxa were represented by 5004 layers that resulted from combining 116 SDM and 102 PGD. This approach showed the highest proportion of area of taxa ranges (on average 41%) and highest representation of PGD within the area of each taxon (on average 76%; Fig. 4; see description in the main text). For some taxa, e.g. Cucurbita pepo, Physalis cinerascens, and Zea mays information on its distribution was assessed at subspecies level rather than at species level, explaining the difference in numbers of CWR taxa.In addition, we included occurrence data of 98 taxa without SDM to prevent missing important areas of taxa known distribution that are important to conserve (see Supplementary Data 3). We enabled the function ‘species of special interest’ (SSI) of Zonation, and included a SSI feature list file, listing the taxon names, as well as taxon-specific coordinate file for each of the 98 taxa that have been reviewed by the experts of each group. The spatial reference system was World Mercator projection. Occurrence data and SDM are treated similarly in the Zonation analysis, i.e., cells where taxa occur will be retained in the solution as long as possible to maximize its representation in the solution.We assigned weights to the 116 taxa with SDM by using IUCN threat categories (according to Supplementary Data 3), giving highest values to taxa with highest risk of extinction that urgently need management actions to further avoid genetic erosion. By including conservation feature weights, Zonation estimates the conservation value of a cell not only based on the presences of a taxa and their distribution range, but also on the weight. A high weight indicates a high conservation value of cells where these taxa are distributed. As there is no rule for weight setting, we assigned values between 1 and 0 regardless of taxa distribution ranges, which is automatically considered in the Zonation algorithm to guarantee the representation of locations where limited-range distributed taxa occur within the most valuable conservation area. Thus, weights were assigned as follows: Critically endangered, CR: 1; Endangered, EN: 1; Vulnerable, VU: 0.8; Near threatened, NT: 0.5; Data deficient, DD: 0.3; Least concern, LC: 0.2 Not evaluated, NE: 0.1. SSI taxa were all weighted similarly with 1 in order to represent the 98 SSI taxa and their occurrences in the top fraction of the most valuable conservation area, as these areas could be considered as ‘irreplaceable’ in terms of conservation. The conservation of these taxa that are only known in a few locations is crucial to maintain their populations. Information on weights for taxa with and without SDM is included in the file that lists the 5004 conservation features and the SSI file, respectively.To include the information on habitat, we included 116 habitat maps which guide the selection of cells to areas where its presence is more probable (see the Methods section: “Habitat preference”). This option can only be used for taxa represented by a raster layer, and is not available for SSI taxa included via occurrence records. By enabling the “landscape condition” option of Zonation, each habitat map is linked to a specific conservation feature layer. Areas with unfavorable habitats will quickly be masked out during the selection of cells in order to obtain a solution that favors conservation areas within areas of preferred habitat.We generated three final scenarios to identify conservation areas for (a) all taxa, (b) taxa exclusively distributing in natural vegetation, and (c) taxa associated with a wider range of habitats such as natural vegetation, agricultural and urban areas. The Zonation configuration remained similar among the three scenarios. When taxa were not included in a given scenario, we assigned a value weight of 0. This excluded the feature to be considered for the hierarchical prioritization of the landscape, but still allowed to evaluate the taxa during post-processing.To evaluate the spatial results (Supplementary Fig. 11), we analyzed performance curves to represent proxies of genetic differentiation within each taxon range (Supplementary Fig. 12). Also, we considered the most valuable 20% area of Mexico to calculate the coincidence of the three scenarios (Supplementary Fig. 13), and the overlap with federal protected areas125 and indigenous regions126,127 (Supplementary Fig. 14), and land cover data used in the analyses (Supplementary Figs. 9, 15).We discussed the proposed methodological framework, input layer and criteria during a fourth workshop in Mexico. It is worth mentioning that we ran several analyses including additional layers, such as areas where indigenous communities live that promote the presence of CWR in the landscape6. However, as the output indicated no evident difference by including this information, final analyses did not consider these data. We neither included protected areas nor tried to expand on the current 12% protected area system, because most management plans do not specifically address CWR management (but see the management program of the Protected Area of ‘Sierra de Manantlán’128), and thus generally do not adequately plan for wild and native genetic resources129. We also discussed different approaches to consider connectivity for taxa, habitats and proxies of genetic differentiation in the Zonation processing. Still, we finally decided to run the analysis without particularly accounting for connectivity as we had no taxa-specific information on dispersal abilities or possible effects of fragmentation, and we did not want to lose efficiency of the solution to represent taxa by or include lower-quality habitats by forcing the solution to an aggregation of pixels.Reporting summaryFurther information on research design is available in the Nature Research Reporting Summary linked to this article. More

Blatchford, M. L., Mannaerts, C. M., Zeng, Y., Nouri, H. & Karimi, P. Status of accuracy in remotely sensed and in-situ agricultural water productivity estimates: A review. Remote Sensing of Environment 234, 111413, https://doi.org/10.1016/j.rse.2019.111413 (2019).Article 
ADS 

Google Scholar 
Geerts, S. & Raes, D. Deficit irrigation as an on-farm strategy to maximize crop water productivity in dry areas. Agricultural Water Management 96, 1275–1284, https://doi.org/10.1016/j.agwat.2009.04.009 (2009).Article 

Google Scholar 
Hellegers, P., Soppe, R., Perry, C. & Bastiaanssen, W. Combining remote sensing and economic analysis to support decisions that affect water productivity. Irrigation Science 27, 243–251, https://doi.org/10.1007/s00271-008-0139-7 (2009).Article 

Google Scholar 
Bastiaanssen, W. G. M. & Steduto, P. The water productivity score (WPS) at global and regional level: Methodology and first results from remote sensing measurements of wheat, rice and maize. The Science of the total environment 575, https://doi.org/10.1016/j.scitotenv.2016.09.032 (2017).Seneviratne, S. I. et al. Investigating soil moisture–climate interactions in a changing climate: A review. Earth Science Reviews 99, https://doi.org/10.1016/j.earscirev.2010.02.004 (2010).Hu, X., Shi, L., Lin, L. & Zha, Y. Nonlinear boundaries of land surface temperature–vegetation index space to estimate water deficit index and evaporation fraction. Agricultural and Forest Meteorology 279, https://doi.org/10.1016/j.agrformet.2019.107736 (2019).Bowen, I. S. The Ratio of Heat Losses by Conduction and by Evaporation from any Water Surface. Physical Review 27, 779–787, https://doi.org/10.1103/PhysRev.27.779 (1926).Article 
ADS 
CAS 
MATH 

Google Scholar 
Penman, H. L. Natural evaporation from open water, hare soil and grass. Proceedings of the Royal Society of London. Series A, Mathematical and physical sciences 193, https://doi.org/10.1098/rspa.1948.0037 (1948).Monteith, J. L. Evaporation and environment. The stage and movement of water in living organisms. Symp.soc.exp.biol.the Company of Biologists (1965).Wang, K. & Dickinson, R. E. A review of global terrestrial evapotranspiration: Observation, modeling, climatology, and climatic variability. Reviews of Geophysics 50, https://doi.org/10.1029/2011RG000373 (2012).Bastiaanssen, W. G. et al. A remote sensing surface energy balance algorithm for land (SEBAL) Part 1: Fomulation. Journal of hydrology 212, 213–229, https://doi.org/10.1016/S0022-1694(98)00253-4 (1998).Article 
ADS 

Google Scholar 
Bastiaanssen, W. G. M. et al. A remote sensing surface energy balance algorithm for land (SEBAL) Part 2. Validation. Journal of Hydrology 212, https://doi.org/10.1016/S0022-1694(98)00254-6 (1998).Su, Z. The Surface Energy Balance System (SEBS) for estimation of turbulent heat fluxes. Hydrology and Earth System Science 6, 85–99, https://doi.org/10.5194/hess-6-85-2002 (2002).Article 
ADS 

Google Scholar 
Norman, J. M., Kustas, W. P. & Humes, K. S. Source approach for estimating soil and vegetation energy fluxes in observations of directional radiometric surface temperature. Agricultural and Forest Meteorology 77, https://doi.org/10.1016/0168-1923(95)02265-y (1995).Mu, Q., Heinsch, F. A., Zhao, M. & Running, S. W. Development of a global evapotranspiration algorithm based on MODIS and global meteorology data. Remote Sensing of Environment 111, https://doi.org/10.1016/j.rse.2007.04.015 (2007).Mu, Q., Zhao, M. & Running, S. W. Improvements to a MODIS global terrestrial evapotranspiration algorithm. Remote Sensing of Environment 115, 1781–1800, https://doi.org/10.1016/j.rse.2011.02.019 (2011).Article 
ADS 

Google Scholar 
Fisher, J. B., Tu, K. P. & Baldocchi, D. D. Global estimates of the land–atmosphere water flux based on monthly AVHRR and ISLSCP-II data, validated at 16 FLUXNET sites. Remote Sensing of Environment 112, 901–919, https://doi.org/10.1016/j.rse.2007.06.025 (2008).Article 
ADS 

Google Scholar 
Kim, H. W., Hwang, K., Mu, Q., Lee, S. O. & Choi, M. Validation of MODIS 16 global terrestrial evapotranspiration products in various climates and land cover types in Asia. KSCE Journal of Civil Engineering 16, https://doi.org/10.1007/s12205-012-0006-1 (2012).Velpuri, N. M., Senay, G. B., Singh, R. K., Bohms, S. & Verdin, J. P. A comprehensive evaluation of two MODIS evapotranspiration products over the conterminous United States: Using point and gridded FLUXNET and water balance ET. Remote Sensing of Environment 139, https://doi.org/10.1016/j.rse.2013.07.013 (2013).Jin, X. et al. Estimation of water productivity in winter wheat using the AquaCrop model with field hyperspectral data. Precision Agriculture 19, 1–17, https://doi.org/10.1007/s11119-016-9469-2 (2016).Article 

Google Scholar 
Felix, R., Clement, A., Igor, S. & Oscar, R. Using Low Resolution Satellite Imagery for Yield Prediction and Yield Anomaly Detection. Remote Sensing 5, 1704–1733, https://doi.org/10.3390/rs5041704 (2013).Article 

Google Scholar 
Lu, Y. et al. Assimilation of soil moisture and canopy cover data improves maize simulation using an under-calibrated crop model. Agricultural Water Management 252, https://doi.org/10.1016/j.agwat.2021.106884 (2021).Jin, X., Kumar, L., Li, Z., Feng, H. & Wang, J. A review of data assimilation of remote sensing and crop models. European Journal of Agronomy 92, https://doi.org/10.1016/j.eja.2017.11.002 (2018).Weiss, M., Jacob, F. & Duveiller, G. Remote sensing for agricultural applications: A meta-review. Remote Sensing of Environment 236, https://doi.org/10.1016/j.rse.2019.111402 (2019).Jin, X. et al. Winter wheat yield estimation based on multi-source medium resolution optical and radar imaging data and the AquaCrop model using the particle swarm optimization algorithm. ISPRS Journal of Photogrammetry and Remote Sensing 126, 24–37 (2017).Article 
ADS 

Google Scholar 
Tao, F., Rötter, R. P., Palosuo, T., Díaz-Ambrona, C. G. H. & Schulman, A. H. Contribution of crop model structure, parameters and climate projections to uncertainty in climate change impact assessments. Global Change Biology 24, https://doi.org/10.1111/gcb.14019 (2017).Jin, X. et al. A review of data assimilation of remote sensing and crop models. European Journal of Agronomy 92, 141–152, https://doi.org/10.1016/j.eja.2017.11.002 (2018).Article 

Google Scholar 
Anikó, K. et al. Statistical modelling of crop yield in Central Europe using climate data and remote sensing vegetation indices. Agricultural and Forest Meteorology 260-261, 300–320, https://doi.org/10.1016/j.agrformet.2018.06.009 (2018).Article 

Google Scholar 
Wang, Y., Zhang, Z., Feng, L., Du, Q. & Runge, T. Combining Multi-Source Data and Machine Learning Approaches to Predict Winter Wheat Yield in the Conterminous United States. Remote Sensing 12, 1232, https://doi.org/10.3390/rs12081232 (2020).Article 
ADS 

Google Scholar 
Franz, T. E. et al. The role of topography, soil, and remotely sensed vegetation condition towards predicting crop yield. Field Crops Research 252, https://doi.org/10.1016/j.fcr.2020.107788 (2020).Noland, R. L. et al. Estimating alfalfa yield and nutritive value using remote sensing and air temperature. Field Crops Research 222, 189–196, https://doi.org/10.1016/j.fcr.2018.01.017 (2018).Article 

Google Scholar 
Cao, J., Zhang, Z., Luo, Y., Zhang, L. & Tao, F. Wheat yield predictions at a county and field scale with deep learning, machine learning, and google earth engine. European Journal of Agronomy, 126204, https://doi.org/10.1016/j.eja.2020.126204 (2021).Jacinta, H. & Kerrie, M. Statistical Machine Learning Methods and Remote Sensing for Sustainable Development Goals: A Review. Remote Sensing 10, 1365, https://doi.org/10.3390/rs10091365 (2018).Article 

Google Scholar 
Jin, X., Liu, S., Baret, F., Hemerlé, M. & Comar, A. Estimates of plant density of wheat crops at emergence from very low altitude UAV imagery. Remote Sensing of Environment 198, 105–114, https://doi.org/10.1016/j.rse.2017.06.007 (2017).Article 
ADS 

Google Scholar 
Maimaitijiang, M. et al. Soybean yield prediction from UAV using multimodal data fusion and deep learning. Remote Sensing of Environment 237, 111599, https://doi.org/10.1016/j.rse.2019.111599 (2020).Article 
ADS 

Google Scholar 
Hossein, A., Mohsen, A., Davoud, A., Salehi, S. H. & Soheil, R. Machine Learning Regression Techniques for the Silage Maize Yield Prediction Using Time-Series Images of Landsat 8 OLI. IEEE Journal of Selected Topics in Applied Earth Observations Remote Sensing PP, 1–15, https://doi.org/10.1109/JSTARS.2018.2823361 (2018).Johansen, K. et al. Predicting Biomass and Yield in a Tomato Phenotyping Experiment Using UAV Imagery and Random Forest. Frontiers in Artificial Intelligence 3, 28, https://doi.org/10.3389/frai.2020.00028 (2020).Article 
PubMed 
PubMed Central 

Google Scholar 
Zhang, L., Ding, X., Shen, Y., Wang, Z. & Wang, X. Spatial Heterogeneity and Influencing Factors of Agricultural Water Use Efficiency in China. Resources and Environment in the Yangtze Basin 28, https://doi.org/10.11870/cjlyzyyhj201904008 (2019).Cheng, M. et al. Satellite time series data reveal interannual and seasonal spatiotemporal evapotranspiration patterns in China in response to effect factors. Agric. Water Manage. 255, https://doi.org/10.1016/j.agwat.2021.107046 (2021).Zhou, L. Comprehensive agricultural regionalization in China. (Agricultural Press of China, 1985).Luo, Y., Zhang, Z., Chen, Y., Li, Z. & Tao, F. ChinaCropPhen1km: A high-resolution crop phenological dataset for three staple crops in China during 2000-2015 based on LAI products. Figshare https://doi.org/10.6084/m9.figshare.8313530.v6 (2019).Luo, Y., Zhang, Z., Chen, Y., Li, Z. & Tao, F. ChinaCropPhen1km: a high-resolution crop phenological dataset for three staple crops in China during 2000–2015 based on leaf area index (LAI) products. Earth System Science Data 12, 197–214, https://doi.org/10.5194/essd-12-197-2020 (2020).Article 
ADS 

Google Scholar 
Song, D. Second China Soil Survey. (Chinese Science Press, 1979).Zhang, T., Yang, X., Wang, H., Li, Y. & Ye, Q. Climatic and technological ceilings for Chinese rice stagnation based on yield gaps and yield trend pattern analysis. Global Change Biology 20, 1289–1298, https://doi.org/10.1111/gcb.12428 (2014).Article 
ADS 
PubMed 

Google Scholar 
Chen, Y., Zhang, Z. & Tao, F. Improving regional winter wheat yield estimation through assimilation of phenology and leaf area index from remote sensing data. European Journal of Agronomy 101, 163–173, https://doi.org/10.1016/j.eja.2018.09.006 (2018).Article 

Google Scholar 
Cheng, M. et al. Combining multi-indicators with machine-learning algorithms for maize yield early prediction at the county-level in China. Agricultural and Forest Meteorology 323, https://doi.org/10.1016/j.agrformet.2022.109057 (2022).Amir, J. & Sinclair, T. A model of the temperature and solar-radiation effects on spring wheat growth and yield. Field Crops Research 28, 47–58, https://doi.org/10.1016/0378-4290(91)90073-5 (1991).Article 

Google Scholar 
Prince, S. D., Haskett, J., Steininger, M. & Wright, S. R. Net Primary Production of U.S. Midwest Croplands from Agricultural Harvest Yield Data. Ecological Applications 11, 1194–1205, https://doi.org/10.1890/1051-0761(2001)011[1194:NPPOUS]2.0.CO;2 (2001).Article 

Google Scholar 
Gilardelli, C. et al. Downscaling rice yield simulation at sub-field scale using remotely sensed LAI data. European journal of agronomy 103, 108–116, https://doi.org/10.1016/j.eja.2018.12.003 (2019).Article 

Google Scholar 
Shakoor, R., Hassan, M. Y., Raheem, A. & Wu, Y.-K. Wake effect modeling: A review of wind farm layout optimization using Jensen׳ s model. Renewable and Sustainable Energy Reviews 58, 1048–1059, https://doi.org/10.1016/j.rser.2015.12.229 (2016).Article 

Google Scholar 
Breiman, L. Random Forests. Machine Learning https://doi.org/10.1023/A:1010933404324 (2001).Article 
MATH 

Google Scholar 
Li, L. et al. Crop yield forecasting and associated optimum lead time analysis based on multi-source environmental data across China. Agricultural and Forest Meteorology 308–309, https://doi.org/10.1016/j.agrformet.2021.108558 (2021).Wang, L. A., Zhou, X., Zhu, X., Dong, Z. & Guo, W. Estimation of biomass in wheat using random forest regression algorithm and remote sensing data. The Crop Journal 4, 212–219, https://doi.org/10.1016/j.cj.2016.01.008 (2016).Article 

Google Scholar 
Feng, P. et al. Dynamic wheat yield forecasts are improved by a hybrid approach using a biophysical model and machine learning technique. Agricultural and Forest Meteorology 285-286, 107922, https://doi.org/10.1016/j.agrformet.2020.107922 (2020).Article 
ADS 

Google Scholar 
Lu, F., Sun, Y. & Hou, F. Using UAV Visible Images to Estimate the Soil Moisture of Steppe. Water 12, 2334, https://doi.org/10.3390/w12092334 (2020).Article 
CAS 

Google Scholar 
Wang, S. et al. High spatial resolution monitoring land surface energy, water and CO2 fluxes from an Unmanned Aerial System. Remote Sensing of Environment 229, 14–31, https://doi.org/10.1016/j.rse.2019.03.040 (2019).Article 
ADS 

Google Scholar 
Chen, Y. et al. Comparison of satellite-based evapotranspiration models over terrestrial ecosystems in China. Remote Sensing of Environment 140, 279–293, https://doi.org/10.1016/j.rse.2013.08.045 (2014).Article 
ADS 

Google Scholar 
Peralta, N., Assefa, Y., Du, J., Barden, C. & Ciampitti, I. Mid-Season High-Resolution Satellite Imagery for Forecasting Site-Specific Corn Yield. Remote Sensing 8, 848, https://doi.org/10.3390/rs8100848 (2016).Article 
ADS 

Google Scholar 
Russello, H. Convolutional neural networks for crop yield prediction using satellite images. IBM Center for Advanced Studies (2018).You, J., Li, X., Low, M., Lobell, D. & Ermon, S. in Proceedings of the AAAI Conference on Artificial Intelligence.Moran, P. A. Notes on continuous stochastic phenomena. Biometrika 37, 17–23 (1950).Article 
MathSciNet 
CAS 

Google Scholar 
Imran, M., Stein, A. & Zurita-Milla, R. Using geographically weighted regression kriging for crop yield mapping in West Africa. International Journal of Geographical Information Systems 29, 234–257, https://doi.org/10.1080/13658816.2014.959522 (2015).Article 

Google Scholar 
Harries, K. Extreme spatial variations in crime density in Baltimore County, MD. Geoforum 37, 404–416, https://doi.org/10.1016/j.geoforum.2005.09.004 (2006).Article 

Google Scholar 
Ghulam, A. et al. Remote Sensing Based Spatial Statistics to Document Tropical Rainforest Transition Pathways. Remote Sensing 7, 6257–6279, https://doi.org/10.3390/rs70506257 (2015).Article 
ADS 

Google Scholar 
Maimaitijiang, M., Ghulam, A., Sandoval, J. S. O. & Maimaitiyiming, M. Drivers of land cover and land use changes in St. Louis metropolitan area over the past 40 years characterized by remote sensing and census population data. International Journal of Applied Earth Observation Geoinformation 35, 161–174, https://doi.org/10.1016/j.jag.2014.08.020 (2015).Article 
ADS 

Google Scholar 
Cheng, M. Long time series (2001-2015) high-resolution crop yield and water productivity dataset of China, Zenodo, https://doi.org/10.5281/zenodo.5121842 (2021).Martens, B., Miralles, D. G., Lievens, H., Schalie, R. D. & Verhoest, N. GLEAM v3: Satellite-based land evaporation and root-zone soil moisture. Geoscientific Model Development 10, https://doi.org/10.5194/gmd-10-1903-2017 (2017).Wang, W., Cui, W., Wang, X. & Chen, X. Evaluation of GLDAS-1 and GLDAS-2 forcing data and Noah model simulations over China at the monthly scale. Journal of Hydrometeorology 17, 2815–2833, https://doi.org/10.1175/JHM-D-15-0191.1 (2016).Article 
ADS 

Google Scholar 
Chen, X. et al. Development of a 10-year (2001–2010) 0.1° data set of land-surface energy balance for mainland China. Atmospheric Chemistry and Physics 14, 14471–14518, https://doi.org/10.5194/acp-14-13097-2014 (2014).Article 
ADS 
CAS 

Google Scholar 
Ramoelo, A. et al. Validation of Global Evapotranspiration Product (MOD16) using Flux Tower Data in the African Savanna, South Africa. Remote Sensing 6, https://doi.org/10.3390/rs6087406 (2014).Yang, X., Yong, B., Ren, L., Zhang, Y. & Long, D. Multi-scale validation of GLEAM evapotranspiration products over China via ChinaFLUX ET measurements. International Journal of Remote Sensing https://doi.org/10.1080/01431161.2017.1346400 (2017).Article 

Google Scholar 
Hu, G., Jia, L. & Menenti, M. Comparison of MOD16 and LSA-SAF MSG evapotranspiration products over Europe for 2011. Remote Sensing of Environment 156, 510–526, https://doi.org/10.1016/j.rse.2014.10.017 (2015).Article 
ADS 

Google Scholar 
Khan, M. S., Liaqat, U. W., Baik, J. & Choi, M. Stand-alone uncertainty characterization of GLEAM, GLDAS and MOD16 evapotranspiration products using an extended triple collocation approach. Agricultural and Forest Meteorology 252, 256–268, https://doi.org/10.1016/j.agrformet.2018.01.022 (2018).Article 
ADS 

Google Scholar 
Glenn, E. P. et al. Scaling sap flux measurements of grazed and ungrazed shrub communities with fine and coarse-resolution remote sensing. Ecohydrology 1, 316–329, https://doi.org/10.1002/eco.19 (2008).Article 

Google Scholar 
Gamon, J. A. Reviews and Syntheses: optical sampling of the flux tower footprint. Biogeosciences 12, 4509–4523, https://doi.org/10.5194/bg-12-4509-2015 (2015).Article 
ADS 

Google Scholar 
Cai, Y. et al. Integrating satellite and climate data to predict wheat yield in Australia using machine learning approaches. Agricultural and Forest Meteorology 274, 144–159, https://doi.org/10.1016/j.agrformet.2019.03.010 (2019).Article 
ADS 

Google Scholar 
Chen, X. et al. Prediction of Maize Yield at the City Level in China Using Multi-Source Data. Remote Sensing 13, https://doi.org/10.3390/rs13010146 (2021).Guo, Y. et al. Integrated phenology and climate in rice yields prediction using machine learning methods. Ecological Indicators 120, 106935, https://doi.org/10.1016/j.ecolind.2020.106935 (2021).Article 

Google Scholar 
Yuan, W. et al. Estimating crop yield using a satellite-based light use efficiency model. Ecological Indicators 60, 702–709, https://doi.org/10.1016/j.ecolind.2015.08.013 (2016).Article 

Google Scholar 
Anandhi, A. Growing degree days – Ecosystem indicator for changing diurnal temperatures and their impact on corn growth stages in Kansas. Ecological Indicators 61, 149–158, https://doi.org/10.1016/j.ecolind.2015.08.023 (2016).Article 

Google Scholar 
Wart, J. V. Estimating Crop Yield Potential At National Scales. Field Crops Research 143, 34–43, https://doi.org/10.1016/j.fcr.2012.11.018 (2013).Article 

Google Scholar 
Kang, Y. S. et al. Yield prediction and validation of onion (Allium cepa L.) using key variables in narrowband hyperspectral imagery and effective accumulated temperature. Computers and Electronics in Agriculture 178, https://doi.org/10.1016/j.compag.2020.105667 (2020).Long, D., Singh, V. P. & Li, Z.-L. How sensitive is SEBAL to changes in input variables, domain size and satellite sensor? Journal of Geophysical Research: Atmospheres 116, https://doi.org/10.1029/2011jd016542 (2011).Liu, Z., Wang, L. & Wang, S. Comparison of Different GPP Models in China Using MODIS Image and ChinaFLUX Data. Remote Sensing 6, 10215–10231, https://doi.org/10.3390/rs61010215 (2014).Article 
ADS 

Google Scholar 
Edreira, J., Guilpart, N., Sadras, V., Cassman, K. G. & Grassini, P. Water productivity of rainfed maize and wheat: A local to global perspective. Agricultural and Forest Meteorology 259, 364–373, https://doi.org/10.1016/j.agrformet.2018.05.019 (2018).Article 
ADS 

Google Scholar 
Li, H. et al. Water Use Characteristics of Maize-Green Manure Intercropping Under Different Nitrogen Application Levels in the Oasis Irrigation Area Scientia Agricultura Sinica 54, 2608–2618 (2021).
Google Scholar 
Wang, S., Ibrom, A., Bauer-Gottwein, P. & Garcia, M. Incorporating diffuse radiation into a light use efficiency and evapotranspiration model: An 11-year study in a high latitude deciduous forest. Agricultural and Forest Meteorology https://doi.org/10.1016/j.agrformet.2017.10.023 (2018).Article 

Google Scholar 
Cheng, M. High-resolution crop yield and water productivity dataset generated using random forest and remote sensing. Zenodo https://doi.org/10.5281/zenodo.6444614 (2022). More