Philippa Kaur

1.
Watt, A. S. Pattern and process in plant community. J. Ecol. 35, 1–22 (1947).
Article  Google Scholar 
2.
Ozinga, W. et al. Local above-ground persistence of vascular plants: Life-history trade-offs and environmental constraints. J. Veg. Sci. 18, 489–497 (2007).
Article  Google Scholar 

3.
Økland, R. H. & Eilertsen, O. Dynamics of understory vegetation in an old-growth boreal coniferous forest, 1988–1993. J. Veg. Sci. 7, 747–762 (1996).
Article  Google Scholar 

4.
Nygaard, P. H. & Ødegaard, T. Sixty years of vegetation dynamics in a south boreal coniferous forest in southern Norway. J. Veg. Sci. 10, 5–16 (1999).
Article  Google Scholar 

5.
Palmer, M. W. & Rusch, G. M. How fast is the carousel? Direct indices of species mobility with examples from an Oklahoma grassland. J. Veg. Sci. 12, 305–318 (2001).
Article  Google Scholar 

6.
Zobel, M., Moora, M. & Herben, T. Clonal mobility and its implications for spatio-temporal patterns of plant communities: What do we need to know next?. Oikos 119, 802–806 (2010).
Article  Google Scholar 

7.
Chaideftou, E., Kallimanis, A. S., Bergmeier, E. & Dimopoulos, P. How does plant species composition change from year to year? A case study from the herbaceous layer of a submediterranean oak woodland. Comm. Ecol. 13, 88–96 (2012).
Article  Google Scholar 

8.
Chapman, J. I. & McEwan, R. W. Spatiotemporal dynamics of α-and β-diversity across topographic gradients in the herbaceous layer of an old-growth deciduous forest. Oikos 122, 1679–1686 (2013).
Article  Google Scholar 

9.
Graae, B. J. & Sunde, P. B. The impact of forest continuity and management on forest floor vegetation evaluated by species traits. Ecography 23, 720–730 (2000).
Article  Google Scholar 

10.
Bakker, J. P., Olff, H., Willems, J. H. & Zobel, M. Why do we need permanent plots in the study of long-term vegetation dynamics?. J. Veg. Sci. 7, 147–156 (1996).
Article  Google Scholar 

11.
Van der Maarel, E. Pattern and process in the plant community: fifty years after A.S. Watt. J. Veg. Sci. 7, 19–28 (1996).
Article  Google Scholar 

12.
Herben, T., Krahulec, F., Hadincová, V. & Skálová, H. Small-scale variability as a mechanism for large-scale stability in mountain grasslands. J. Veg. Sci. 4, 163–170 (1993).
Article  Google Scholar 

13.
Økland, R. H. Persistence of vascular plants in a Norwegian boreal coniferous forest. Ecography 18, 3–14 (1995).
Article  Google Scholar 

14.
Campetella, G. et al. Patterns of plant trait-environment relationship along a forest succession chronosequence. Agric. Ecosyst. Environ. 145, 38–48 (2011).
Article  Google Scholar 

15.
Canullo, R. et al. Patterns of clonal growth modes along a chronosequence of post-coppice forest regeneration in beech forest of Central Italy. Fol. Geobot. 46, 271–288 (2011).
Article  Google Scholar 

16.
Rūsiņa, S., Gavrilova, I., Roze, I. & Šulcs, V. Temporal species turnover and plant community changes across different habitats in the lake Engure nature park Latvia. Proc. Latv. Acad. Sci. Sect. B. Nat. Exact Appl. Sci. 68, 68–79 (2014).
Google Scholar 

17.
Norden, B. & Appelqvist, T. Conceptual problems of ecological continuity and its bioindicators. Biodivers. Conserv. 10, 779–791 (2001).
Article  Google Scholar 

18.
Bartha, S., Canullo, R., Chelli, S. & Campetella, G. Unimodal relationships of understory alpha and beta diversity along chronosequence in coppiced and unmanaged beech forests. Diversity 12, 101 (2020).
Article  Google Scholar 

19.
Gilliam, F. S. The ecological significance of the herbaceous layer in temperate forest ecosystems. Bioscience 57, 845–857 (2007).
Article  Google Scholar 

20.
Campetella, G. et al. Scale dependent effects of coppicing on the species pool of late successional beech forest in the Central Apennines (Italy). Appl. Veg. Sci. 19, 474–485 (2016).
Article  Google Scholar 

21.
Lavorel, S. & Garnier, E. Predicting changes in community composition and ecosystem functioning from plant traits: Revisiting the Holy Grail. Funct. Ecol. 16, 545–556 (2002).
Article  Google Scholar 

22.
Weiher, E. et al. Challenging theophrastus: A common core list of plant traits for functional ecology. J. Veg. Sci. 10, 609–620 (1999).
Article  Google Scholar 

23.
Westoby, M. A Leaf-height-seed (LHS) plant ecology strategy scheme. Plant Soil 199, 213–227 (1998).
CAS  Article  Google Scholar 

24.
Wright, I. J. et al. The worldwide leaf economic spectrum. Nature 428, 821–827 (2004).
ADS  CAS  Article  Google Scholar 

25.
Klimešová, J., Martínková, J. & Ottaviani, G. Belowground plant functional ecology: Towards an integrated perspective. Funct. Ecol. 32, 2115–2126 (2018).
Article  Google Scholar 

26.
de Bello, F. et al. On the need for phylogenetic ‘corrections’ in functional trait-based approaches. Fol. Geobot. 50, 349–357 (2015).
Article  Google Scholar 

27.
Aubin, I., Messier, C. & Bouchard, A. Can plantations develop understory biological and physical attributes of naturally regenerated forests?. Biol. Conserv. 141, 2462–2476 (2008).
Article  Google Scholar 

28.
Dahlgren, J. P., Eriksson, O., Bolmgren, K., Strindell, M. & Ehrlén, J. Specific leaf area as a superior predictor of changes in field layer abundance during forest succession. J. Veg. Sci. 17, 577–582 (2006).
Article  Google Scholar 

29.
Wellstein, C. et al. Effects of extreme drought on specific leaf area of grassland species: A meta-analysis of experimental studies in temperate and sub-Mediterranean systems. Glob. Change Biol. 23, 2473–2481 (2017).
ADS  Article  Google Scholar 

30.
Lindacher, R., Böcker, R., Bemmerlein-Lux, F. A., Kleemann, A. & Haas, S. PHANART Datenbank der Gefäßpflanzen Mitteleuropas, Erklärung der Kennzahlen, Aufbau und Inhalt. Veröff. Geobot. Inst. ETH, Stift. Rübel 125, 1–436 (1995).
Google Scholar 

31.
Turner, I. M. Sclerophylly: Primarily protective?. Funct. Ecol. 8, 669–675 (1994).
Article  Google Scholar 

32.
Van Groenendael, J. M., Klimeš, L., Klimešová, J. & Hendriks, R. J. J. Comparative ecology of clonal plants. Philos. Trans. Roy. Soc. B 351, 1331–1339 (1996).
ADS  Article  Google Scholar 

33.
Sammul, M., Kull, K., Niitla, T. & Mols, T. A comparison of plant communities on the basis of their clonal growth patterns. Evol. Ecol. 18, 443–467 (2004).
Article  Google Scholar 

34.
Canullo, R. et al. Unravelling mechanisms of short-term vegetation dynamics in complex coppice forest systems. Fol. Geobot. 52, 71–81 (2017).
Article  Google Scholar 

35.
Kidson, R. & Westoby, M. Seed mass and seedling dimensions in relation to seedling establishment. Oecologia 125, 11–17 (2000).
ADS  CAS  Article  Google Scholar 

36.
Moles, A. T. & Westoby, M. Seed size and plant strategy across the whole life cycle. Oikos 113, 91–105 (2006).
Article  Google Scholar 

37.
Campetella, G., Canullo, R. & Allegrini, M. C. Status and changes of ground vegetation at the CONECOFOR plots, 1999–2005. Ann. Silvicult. Res. 34, 29–48 (2008).
Google Scholar 

38.
Wright, I. J., Reich, P. B. & Westoby, M. Strategy shifts in leaf physiology, structure and nutrient content between species of high- and low-rainfall and high- and low-nutrient habitats. Funct. Ecol. 15, 423–434 (2001).
Article  Google Scholar 

39.
Ackerly, D. D. Functional traits of chaparral shrubs in relation to seasonal water deficit and disturbance. Ecol. Monogr. 74, 25–44 (2004).
Article  Google Scholar 

40.
Kopecký, M., Hédl, R. & Szabó, P. Non-random extinctions dominate plant community changes in abandoned coppices. J. Appl. Ecol. 50, 79–87 (2013).
Article  Google Scholar 

41.
Naaf, T. & Wulf, M. Traits of winner and loser species indicate drivers of herb layer changes over two decades in forests of NW Germany. J. Veg. Sci. 22, 516–527 (2011).
Article  Google Scholar 

42.
Ottaviani, G., Martínková, J., Herben, T., Pausas, J. G. & Klimešová, J. On plant modularity traits: Functions and challenges. Trends Plant Sci. 22, 648–651 (2017).
CAS  Article  Google Scholar 

43.
Klimešová, J. & Klimeš, L. Bud banks and their role in vegetative regeneration—A literature review and proposal for simple classification and assessment. Perspect. Plant Ecol. Evol. Syst. 8, 115–129 (2007).
Article  Google Scholar 

44.
Chelli, S. et al. Climate is the main driver of clonal and bud bank traits in Italian forest understories. Persp. Plant Ecol. Evol. Syst. 40, 125478 (2019).
Article  Google Scholar 

45.
Grime, J. P. Benefits of plant diversity to ecosystems: Immediate, filter and founder effects. J. Ecol. 86, 902–910 (1998).
Article  Google Scholar 

46.
Alpert, P. & Simms, E. L. The relative advantages of plasticity and fixity in different environments: When is it good for a plant to adjust?. Evol. Ecol. 16, 285–297 (2002).
Article  Google Scholar 

47.
Denney, D. A., Jameel, M. I., Bemmels, J. B., Rochford, M. E. & Anderson, J. T. Small spaces, big impacts: Contributions of micro-environmental variation to population persistence under climate change. AoB Plants 12, 5 (2020).
Article  Google Scholar 

48.
Swenson, N. G. et al. Temporal turnover in the composition of tropical tree communities: Functional determinism and phylogenetic stochasticity. Ecology 93, 490–499 (2012).
Article  Google Scholar 

49.
Kichenin, E., Wardle, D. A., Peltzer, D. A., Morse, C. W. & Freschet, G. T. Contrasting effects of plant inter-and intraspecific variation on community-level trait measures along an environmental gradient. Funct. Ecol. 27, 1254–1261 (2013).
Article  Google Scholar 

50.
Petriccione, B. & Pompei, E. The CONECOFOR programme: general presentation, aims and co-ordination. J. Limnol. 61, 3–11 (2002).
Article  Google Scholar 

51.
Hijmans, R. J., Cameron, S. E., Parra, J. L., Jones, P. G. & Jarvis, A. Very high resolution interpolated climate surfaces for global land areas. Int. J. Climatol. 25, 1965–1978 (2005).
Article  Google Scholar 

52.
Bagnouls, F. & Gaussen, H. Les climats biologiques et leur classification. Ann. Geogr. 335, 193–220 (1957).
Article  Google Scholar 

53.
FAO/UNESCO/WMO. World map of desertification. Food and Agricultural, Organization, Rome (1997).

54.
EUFORGEN. Distribution map of Beech (Fagus sylvatica), www.euforgen.org (2009).

55.
Dupouey, J. L. Manual on methods and criteria for harmonized sampling, assessment, monitoring and analysis of the effects of air pollution on forests Part VIII. Assessment of Ground Vegetation (ICP-Forests, Hamburg, 1998).
Google Scholar 

56.
Canullo, R., Campetella, G., Allegrini, M. C. & Smargiassi, V. Management of forest vegetation data series: The role of database in the frame of quality assurance procedure. J. Limnol. 61, 100–105 (2002).
Article  Google Scholar 

57.
Klimeš, L., Klimešová, J., Hendriks, R. & van Groenendael, J. Clonal plant architectures: a comparative analysis of form and function. In The Ecology and Evolution of Clonal Plants (eds de Kroon, H. & van Groenendael, J.) 1–29 (Backhuys Publishers, Leiden, 1997).
Google Scholar 

58.
Cerabolini, B., Ceriani, R. M., Caccianiga, M., De Andreis, R. & Raimondi, B. Seed size, shape and persistence in soil: A test on Italian flora from Alps to Mediterranean coasts. Seed Sci. Res. 13, 75–85 (2003).
Article  Google Scholar 

59.
Royal Botanical Gardens Kew. Seed Information Database (SID), Version 7.1. Available from https://data.kew.org/sid/ (2008).

60.
Kleyer, M. et al. The LEDA Traitbase: A database of plant life-history traits of North West European Flora. J. Ecol. 96, 1266–1274 (2008).
Article  Google Scholar 

61.
Wellstein, C. & Kuss, P. Diversity and frequency of clonal traits along natural and land-use gradients in grasslands of the Swiss Alps. Fol. Geobot. 46, 255–270 (2011).
Article  Google Scholar 

62.
Pérez-Harguindeguy, N. et al. New handbook for standardised measurement of plant functional traits worldwide. Austr. J. Bot. 61, 167–234 (2013).
Article  Google Scholar 

63.
Reinecke, J., Klemm, G. & Heinken, T. Vegetation change and homogenization of species composition in temperate nutrient deficient Scots pine forests after 45 yr. J. Veg. Sci. 25, 113–121 (2014).
Article  Google Scholar 

64.
Anderson, M. J. A new method for non-parametric multivariate analysis of variance. Aust. Ecol. 26, 32–46 (2001).
Google Scholar 

65.
Saitou, N. & Nei, M. The neighbor-joining method: A new method for reconstructing phylogenetic trees. Mol. Biol. Evol. 4, 406–425 (1987).
CAS  PubMed  Google Scholar 

66.
Tamura, K., Nei, M. & Kumar, S. Prospects for inferring very large phylogenies by using the neighbor-joining method. Proc. Nat. Acad. Sci. USA 101, 11030–11035 (2004).
ADS  CAS  Article  Google Scholar 

67.
Kumar, S., Stecher, G., Li, M., Knyaz, C. & Tamura, K. MEGA X: Molecular evolutionary genetics analysis across computing platforms. Mol. Biol. Evol. 35, 1547–1549 (2018).
CAS  Article  Google Scholar 

68.
Weiher, E., Clarke, G. D. P. & Keddy, P. A. Community assembly rules, morphological dispersion, and the coexistence of plant species. Oikos 81, 309 (1998).
Article  Google Scholar 

69.
Breiman, L., Friedman, J. H., Olshen, R. A. & Stone, C. J. Classification and regression trees (Wadsworth International Group, Belmont, 1984).
Google Scholar 

70.
Ryo, M. & Rillig, M. C. Statistically reinforced machine learning for nonlinear patterns and variable interactions. Ecosphere 8(11), e01976 (2017).
Article  Google Scholar 

71.
De’ath, G. & Fabricius, K. E. Classification and regression trees: A powerful yet simple technique for ecological data analysis. Ecology 81, 3178–3192 (2000).
Article  Google Scholar 

72.
Oksanen, J., Blanchet, F. G., Kindt, R., Legendre, P., Minchin, P.R., O’Hara, R.B., Simpson, G.L., Solymos, P., Stevens, M. H. H. & Wagner, H. Vegan: Community Ecology Package. R package version 2.0–7. (2013) Available at https://CRAN.R-project.org/package=vegan

73.
Hothorn, T., Hornik, K. & Zeileis, A. Unbiased recursive partitioning: A conditional inference framework. J. Comput. Graph. Stat. 15, 651–674 (2006).
MathSciNet  Article  Google Scholar 

74.
Kembel, S. W. et al. Picante: R tools for integrating phylogenies and ecology. Bioinformatics 26, 1463–1464 (2010).
CAS  Article  Google Scholar 

75.
Laliberté, E. & Legendre, P. A distance-based framework for measuring functional diversity from multiple traits. Ecology 91, 299–305 (2010).
Article  Google Scholar 

76.
Fabbio, G., Manetti, M. C. & Bertini, G. Aspects of biological diversity in the CONECOFOR plots. I. Structural and species diversity in the tree community. Ann. Silvicul. Res. 30, 17–28 (2006).
Google Scholar 

77.
Trabucco, A. & Zomer, R. J. Global aridity index (global-aridity) and global potential evapo-transpiration (global-PET) geospatial database. CGIAR Consortium for Spatial Information (2009). More

Weather
All the weather parameters measured during the study period were similar to the long-term averages (Fig. S1). During the study period (2014–2018), crops received an average annual rainfall of 763 mm, although its distribution was quite different amongst the rainy season (June–September) (Fig. S1). Rice/maize season in 2014, 2015, and 2016, 2017 received 485 (256 mm in September), 420 (255 mm in July), 533 (284 mm in August), and 695 mm (247 mm in June and 226 mm in September) of rainfall, respectively. In 1st year, the wheat crop receivedrainfall of 247 mm whereas in the 2nd, 3rd, and 4th years it was only 56, 96 and 78 mm, respectively.
Crops and system productivity
The management practices under different rice/maize-based scenarios influenced the crop grain yields over the 4-years (2014–2017) (Table 1). Scenarios with rice crops (Sc1-Sc3) did not differ in rice yields during the year 2014 and 2017, but CT direct seeded rice (Sc2) in the 2nd year (2015) and ZT direct seeded rice (Sc3) in the 3rd year (2016) produced 0.9 Mg ha−1 higher and 1.1 Mg ha−1 lower yield than farmers’ practice (Sc1), respectively (Table 1). Rice equivalent maize yields in CA-based scenarios (Sc6-Sc7) did not differ from scenarios with rice crops (Sc1-Sc3) in any of the study years. Rice equivalent maize yield of CA-based Sc5 with maize on PB, although was similar to Sc1 in all the years but was 1.41 Mg ha−1 lower than ZT-DSR (Sc3) in 1st year and 0.98 Mg ha−1 lower than CT-DSR (Sc2) in 2nd year. In contrast, rice equivalent yield (REY) of Sc4 with maize on fresh beds (FB) produced lower yields than one of the rice-based scenarios in three out of four years. These results suggest that maize performs better under CA-based management system than under conventional tillage system. Almost 5% higher yield of maize was recorded in the 1st year and 12–16% higher in the last three years under CA-based scenario (Sc7) compared to CT-based scenario (Sc4) and at par with Sc5. Based on the 4-years average, rice equivalent yield (REY) of Sc4 (maize on FB ) was 0.8 Mg ha−1 (12%) lower than Sc1 (business-as-usual) whereas other scenarios did not differ from each other in REY (Table 1).
Table 1 Effect of different scenarios on grain yields (Mg ha−1) of rice, maize, wheat and systems during 4-years (2014–2018).
Full size table

The management practices influenced wheat grain yield over the years of experimentation (Table 1). Across study years, the grain yield of ZT wheat in CA-based scenario was either similar or higher than CT wheat. Results showed significantly (P  Sc2 = Sc3 (1753–1759 mm ha−1)  > S7 = Sc6 = Sc4 = S5 (289–365 mm ha−1) (Table 2). The same trend followed in all the study years except in the 4th year, where irrigation water input in Sc5 (maize on PB) was 109–154 mm ha−1 (22–28%) lower than Sc6 and Sc7 (ZT maize on flat beds). The amount of water applied in CT-based rice crop (Sc1; farmers’ practice) was significantly (P  Sc2 = Sc3  > Sc7  > Sc4-Sc6.
Higher grain yield and low water use led to significantly (P  Sc4 (2.25 kg grain m−3)  > Sc7 (2.15 kg grain m−3)  > Sc6 (2.06 kg grain m−3), respectively compared to Sc1 (0.30 kg grain m−3) (Table 2). In wheat, CA-based management practices increased WPI by 9% (4-years’ mean) compared to Sc1 (1.21 kg grain m−3). CA-based management practices improved mean WPI by 23 and 438% in RW and MW system, respectively compared to Sc1 (0.42 kg grain m−3).
Figure 3

Effect of different scenarios on irrigation water productivity (kg grain m−3) of rice, maize, wheat and systems during 4-years (2014–2018).

Full size image

Energy use efficiency
Energy equivalents for different agricultural operations used in the study are given in Table S2. The energy input and output (Tables S3 and S4), and energy use efficiency (EUE) of rice, maize, wheat and mungbean were influenced by the management practices and varied from year to year (Fig. 4). During rice/maize, higher EUE was observed in maize based scenarios (Sc4-Sc7) than in rice-based scenarios (Sc1-Sc3) (10.81–13.83 MJ MJ−1 versus 3.95–4.85 MJ MJ−1) (Table 2). Rice-based scenarios (Sc1-Sc3) did not differ in EUE. However, in maize-based scenarios (Sc4-Sc7), EUE of CA-based maize scenarios (Sc5-Sc7) was 17–28% higher than CT-based maize Sc4. Across years also, the same trend was observed with no difference in EUE of rice-based scenarios (Sc1-Sc3), whereas CA-based maize scenarios (Sc5-Sc7) had higher EUE than CT-based Sc4 (Table 2). In wheat crop, highest EUE was observed under CA-based scenarios (Sc2-Sc3 and Sc5-Sc7) compared to CT-based scenarios (Sc1 and Sc4) across all study years and based on four years’ average (9.26–10.05 MJ MJ−1 versus 7.44–7.84 MJ MJ−1), it is indicated that CA-based scenarios are more energy-efficient than those of CT-based scenarios (Fig. 4). In all the years, EUE of maize-based scenarios (Sc4-Sc7) were higher than rice-based scenarios (Sc1-Sc3) but within rice-based scenarios (Sc1-Sc3), results were more variable with higher EUE of CA-based Sc2 and Sc3 in 1st and 2nd year than CT-based scenarios (Sc1) but did not differ in 3rd and 4th year (Fig. 4). On system basis, the EUE of different scenarios decreased in the following order: Sc5 (11.92 MJ MJ−1)  > Sc6 = Sc7 (10.26–10.95 MJ MJ−1)  > Sc4 (9.25 MJ MJ−1)  > Sc3 = Sc2 (6.23–6.25 MJ MJ−1)  > Sc1 (5.05 MJ MJ−1) (Table 2). Maize-based scenarios (Sc5-Sc7) had 48 to 136% higher EUE than rice-based scenarios (Sc1-Sc3) suggesting maize-wheat based cropping systems were more efficient in energy use than rice–wheat based systems (Table 2). Scenario 3 (+ 24%) in RW and Sc5 (+ 136%) in MW system were the most energy-efficient among the different combinations of management practices in 4-years of study.
Figure 4

Effect of different scenarios on energy use efficiency of rice, maize, wheat and systems during 4-years (2014–2018).

Full size image

Methane (CH4) and nitrous oxide (N2O) emission from soil
Methane (CH4) was emitted only from the rice plots (Table 3). The estimated mean value of CH4 emission (kg CO2 eq. ha−1) was 39% lower in CA-based rice scenarios without continuous flooding (Sc2 and Sc3) compared to CT-based Sc1 with continuous flooding for  > 1 month (Table 3).
Table 3 Effect of different scenarios on GHGs emissions, C-sequestration and GWP of rice, maize, wheat and systems (based on 4-year average, 2014–18).
Full size table

N2O emission varied from 7 to 583 kg CO2 eq. ha−1 during the rice season (Table 3). The maximum amount of N2O emission (580–583 kg CO2 eq. ha−1) was observed in CA-based rice scenarios (Sc2-Sc3) followed by the maize-based scenarios (50–61 kg CO2 eq. ha−1) and was the lowest in CT-based rice Sc1 (7 kg CO2 eq. ha−1). The CA-based rice and maize scenarios produced 88 and 9 times higher N2O emission compared to Sc1, respectively. The N2O emission in the wheat season ranged between 50 to 102 kg CO2 eq. ha−1 (Table 3). The highest N2O emission was estimated with CA-based scenarios (Sc2-Sc3) (101–102 kg CO2 eq. ha−1) and followed by scenarios Sc5-Sc7 (72–73 kg CO2 eq. ha−1) and was lowest in CT-based scenarios Sc1 and Sc4 (50 kg CO2 eq. ha−1). The N2O emission in the wheat crop was increased by 57% under CA-based management scenarios compared to CT-based management scenario (Table 3). On system basis, CA-based rice and maize systems emitted 12 and 2.4 times more N2O compared to Sc1, respectively but methane emission was reduced to zero (Table 3). Overall CA-based cereal management systems emitted almost six-time higher N2O emission compared to farmers’ practice, irrespective of cropping systems (Table 3).
GHG emission associated with residue burning (kg CO2 eq. ha−1)
Crop residue burning is a common farmers’ practice in the western IGP. Therefore, GHG emission due to residue burning (kg CO2 eq. ha−1) was estimated with CT-based system of rice (Sc1; 278 kg CO2 eq. ha−1) and maize (Sc4; 69 kg CO2 eq. ha−1) cultivation (Table 3). In the case of wheat, the GHG emission due to residue burning (kg CO2 eq. ha−1) was estimated with CT-based cultivation of wheat in Sc1 (59 kg CO2 eq. ha−1) and Sc4 (40 kg CO2 eq. ha−1). No GHG emission (kg CO2 eq. ha−1) was considered due to burning where crop residues were retained/incorporated in CA-based management practices under different scenarios.
GHG emission due to energy consumption (kg CO2 eq. ha−1)
GHG emission due to energy consumption varied from 2414 to 2941, 1005 to 1126 and 1122 to 1299 kg CO2 eq. ha−1 in rice, maize, and wheat, respectively (Table 3). Compared to CA-based management scenarios, CT-based scenarios emitted more GHGs due to the higher consumption of electricity and diesel energy in all the crops and cropping systems. Compared to Sc1, GHG emission due to energy consumption from rice/maize season was 16–18% lower in CA-based rice scenarios (Sc2-Sc3) and 63–66% lower in maize-based scenarios (Sc4-Sc7) (Table 3). Overall, compared to Sc1, CA-based scenarios reduced ~ 17 and 63% of GHG emissions due to energy consumption in rice and maize across the years, respectively. Similarly, in wheat, CA-based scenarios (Sc2-Sc3 and Sc5-Sc7) reduced 10% GHG emission due to energy consumptions as compared to CT-based scenarios (Sc1 and Sc4). On the system basis, Sc2, Sc3, Sc4, Sc5, Sc6, and Sc7 recorded lower energy-related emission of GHG by 14, 15, 43, 50, 46, and 43% (4-years’ mean), respectively, relative to Sc1 (4240 kg CO2 eq. ha−1) (Table 3). Rice and maize-based systems recorded ~ 15 and 46% lower GHG related emissions, respectively compared to farmers’ practice (Sc1-4240 kg CO2 eq. ha−1).
Carbon (C) sequestration
The estimated C-sequestration was carried out in those scenarios where crop residues were retained/ incorporated during the study period. The C-sequestration varied with the amount of crop residue was recycled under different crops and cropping systems. Estimated C-sequestration in soil varied from 0 to − 625 kg CO2 eq. ha−1 in rice, 0 to − 908 CO2 eq. ha−1 in maize and 0 to − 1821 kg CO2 eq. ha−1 in wheat (Table 3). On system basis, the highest C-sequestration was estimated under CA-based management scenarios which varied in the following order of Sc7 (3039 kg CO2 eq. ha−1)  > Sc3 (2446 kg CO2 eq. ha−1)  > Sc2 (2086 kg CO2 ha−1)  > Sc6 (2070 kg CO2 eq. ha−1).
Total global warming potential (GWP)
Global warming potential (GWP) varied with crop management practices under different scenarios over the years. In 4-year, the total estimated GWP from rice was lower under the CA-based systems than CT-based system. On 4-year mean basis, the GWP under the CA-based rice (Sc2-Sc3) and maize (Sc5-Sc7) systems were lowered by ~ 28 and 90% compared to farmers’ practice (Sc1), respectively (Table 3). Within maize-based scenarios, the CA-based scenarios (Sc5-Sc7) reduced the GWP of maize by 77–83% compared to CT-based Sc4. The GWP in wheat varied from − 384 to 1409 kg CO2 eq. ha−1 based on 4 year average (Table 3). The 4 years mean GWP was significantly lower by 127–138% in CA-based RW system (Sc2-Sc3) and 96–99% in CA-based MW system (Sc5-Sc7) compared to Sc1, respectively (Table 3). The mean GWP of wheat under CT-based RW system (Sc1) was similar to CT-based MW (Sc1and Sc4) systems.
The crop management practices under different scenarios influenced the total GWP (CO2 eq. ha−1) in both the cropping systems (RW and MW system) during the study years (Table 3). On 4-years system mean basis, GWP under Sc2, Sc3, Sc4, Sc5, Sc6, and Sc7 were 48, 54, 59, 96, 95, and 107% lower compared to Sc1 (farmers’ practice), respectively. In CA-based RW and MW systems, GWP was estimated lower by 50 and 89% compared to CT-based Sc1(6451 kg CO2 eq. ha−1), respectively. More

1.
Zhang, Y. et al. Concentrations and chemical compositions of fine particles (PM2.5) during haze and non-haze days in Beijing. Atmos. Res. 174–175, 62–69 (2016).
Article  CAS  Google Scholar 
2.
Li, Y. et al. Observations of ammonia, nitric acid, and fine particles in a rural gas production region. Atmos. Environ. 83, 80–89 (2014).
ADS  CAS  Article  Google Scholar 

3.
Singh, J., Kunhikrishnan, A., Bolan, N. S. & Saggar, S. Impact of urease inhibitor on ammonia and nitrous oxide emissions from temperate pasture soil cores receiving urea fertilizer and cattle urine. Sci. Total Environ. 465, 56–63 (2013).
ADS  CAS  Article  Google Scholar 

4.
Sha, Z., Li, Q., Lv, T., Misselbrook, T. & Liu, X. Response of ammonia volatilization to biochar addition: A meta-analysis. Sci. Total Environ. 655, 1387–1396 (2019).
ADS  CAS  Article  Google Scholar 

5.
Wang, H. et al. Ammonia emissions from paddy fields are underestimated in China. Environ. Pollut. 235, 482–488 (2018).
CAS  Article  Google Scholar 

6.
Paustian, K. et al. Climate-smart soils. Nature 532, 49–57 (2016).
ADS  CAS  Article  Google Scholar 

7.
Huang, Y. et al. Methane and nitrous oxide flux after biochar application in subtropical acidic paddy soils under tobacco-rice rotation. Sci. Rep. 9, 17277 (2019).
ADS  Article  CAS  Google Scholar 

8.
Sheng, Y. & Zhu, L. Biochar alters microbial community and carbon sequestration potential across different soil pH. Sci. Total Environ. 622–623, 1391–1399 (2018).
ADS  Article  CAS  Google Scholar 

9.
Liu, Q. et al. How does biochar influence soil N cycle? A meta-analysis. Plant Soil 426, 211–225 (2018).
CAS  Article  Google Scholar 

10.
Huang, M., Fan, L., Chen, J., Jiang, L. & Zou, Y. Continuous applications of biochar to rice: Effects on nitrogen uptake and utilization. Sci. Rep. 8, 11461 (2018).
ADS  Article  CAS  Google Scholar 

11.
Feng, Y. et al. Biochar applied at an appropriate rate can avoid increasing NH3 volatilization dramatically in rice paddy soil. Chemosphere 168, 1277–1284 (2016).
ADS  Article  CAS  Google Scholar 

12.
Sun, H., Zhang, H., Min, J., Feng, Y. & Shi, W. Controlled-release fertilizer, floating duckweed, and biochar affect ammonia volatilization and nitrous oxide emission from rice paddy fields irrigated with nitrogen-rich wastewater. Paddy Water Environ. 14, 1–7 (2015).
Google Scholar 

13.
Sun, H., Lu, H., Chu, L., Shao, H. & Shi, W. Biochar applied with appropriate rates can reduce N leaching, keep N retention and not increase NH3 volatilization in a coastal saline soil. Sci. Total Environ. 575, 820–825 (2017).
ADS  CAS  Article  Google Scholar 

14.
Sun, H. et al. Responses of ammonia volatilization from rice paddy soil to application of wood vinegar alone or combined with biochar. Chemosphere 242, 125247 (2020).
ADS  CAS  Article  Google Scholar 

15.
Nizamuddin, S. et al. Upgradation of chemical, fuel, thermal, and structural properties of rice husk through microwave-assisted hydrothermal carbonization. Environ. Sci Pollut. R. 25, 17529–17539 (2018).
CAS  Article  Google Scholar 

16.
Gronwald, M., Vos, C., Helfrich, M. & Don, A. Stability of pyrochar and hydrochar in agricultural soil—A new field incubation method. Geoderma 284, 85–92 (2016).
ADS  CAS  Article  Google Scholar 

17.
Liu, Y. et al. Effect of pyrochar and hydrochar on water evaporation in clayey soil under greenhouse cultivation. Environ. Res. Public Health 16, 2580 (2019).
CAS  Article  Google Scholar 

18.
Malghani, S., Gleixner, G. & Trumbore, S. E. Chars produced by slow pyrolysis and hydrothermal carbonization vary in carbon sequestration potential and greenhouse gases emissions. Soil Biol. Biochem. 62, 137–146 (2013).
CAS  Article  Google Scholar 

19.
Han, L. et al. New evidence for high sorption capacity of hydrochar for hydrophobic organic pollutants. Environ. Sci. Technol. 50, 13274–13282 (2016).
ADS  CAS  Article  Google Scholar 

20.
Hua, Y. et al. Microbial aging of hydrochar as a way to increase cadmium ion adsorption capacity: Process and mechanism. Bioresour. Technol. 300, 122708 (2020).
CAS  Article  Google Scholar 

21.
Schimmelpfennig, S., Müller, C., Grünhage, L., Koch, C. & Kammann, C. Biochar, hydrochar and uncarbonized feedstock application to permanent grassland—Effects on greenhouse gas emissions and plant growth. Agric. Ecosyst. Environ. 191, 39–52 (2014).
CAS  Article  Google Scholar 

22.
Elaigwu, S. E. & Greenway, G. M. Microwave-assisted and conventional hydrothermal carbonization of lignocellulosic waste material: Comparison of the chemical and structural properties of the hydrochars. J. Anal. Appl. Pyrol. 118, 1–8 (2016).
CAS  Article  Google Scholar 

23.
Huang, R. & Tang, Y. Speciation dynamics of phosphorus during (hydro)thermal treatments of sewage sludge. Environ. Sci. Technol. 49, 14466–14474 (2015).
ADS  CAS  Article  Google Scholar 

24.
Donar, Y. O., Çağlar, E. & Sınağ, A. Preparation and characterization of agricultural waste biomass based hydrochars. Fuel 183, 366–372 (2016).
CAS  Article  Google Scholar 

25.
Mumme, J. et al. Hydrothermal carbonization of digestate in the presence of zeolite: Process efficiency and composite properties. ACS Sustain. Chem. Eng. 3, 2967–2974 (2015).
CAS  Article  Google Scholar 

26.
Bargmann, I., Rillig, M., Kruse, A., Greef, J. & Kücke, M. Effects of hydrochar application on the dynamics of soluble nitrogen in soils and on plant availability. J. Plant. Nutr. Soil. Sci. 177, 48–58 (2014).
Article  CAS  Google Scholar 

27.
Chu, Q. et al. Microalgae-derived hydrochar application on rice paddy soil: Higher rice yield but increased gaseous nitrogen loss. Sci. Total Environ. 717, 137127 (2020).
ADS  CAS  Article  Google Scholar 

28.
Chu, Q. et al. Sewage sludge-derived hydrochar that inhibits ammonia volatilization, improves soil nitrogen retention and rice nitrogen utilization. Chemosphere 245, 125558 (2020).
ADS  CAS  Article  Google Scholar 

29.
Xue, Y. et al. Hydrogen peroxide modification enhances the ability of biochar (hydrochar) produced from hydrothermal carbonization of peanut hull to remove aqueous heavy metals: Batch and column tests. Chem. Eng. J. 200–202, 673–680 (2012).
Article  CAS  Google Scholar 

30.
Yu, S. et al. Biowaste to treasure: Application of microbial-aged hydrochar in rice paddy could improve nitrogen use efficiency and rice grain free amino acids. J. Clean Prod. 240, 118180 (2019).
CAS  Article  Google Scholar 

31.
Ti, C., Xia, L., Chang, S. X. & Yan, X. Potential for mitigating global agricultural ammonia emission: A meta-analysis. Environ. Pollut. 245, 141–148 (2019).
CAS  Article  Google Scholar 

32.
Chu, Q. et al. Bentonite hydrochar composites mitigate ammonia volatilization from paddy soil and improve nitrogen use efficiency. Sci. Total Environ. 718, 137301 (2020).
ADS  CAS  Article  Google Scholar 

33.
Luo, S. et al. Long-term biochar application influences soil microbial community and its potential roles in semiarid farmland. Appl. Soil Ecol. 117–118, 10–15 (2017).
Article  Google Scholar 

34.
Zhu, X., Chen, B., Zhu, L. & Xing, B. Effects and mechanisms of biochar-microbe interactions in soil improvement and pollution remediation: A review. Environ. Pollut. 227, 98–115 (2017).
CAS  Article  Google Scholar 

35.
Rutherford, D. W., Wershaw, R. L., Rostad, C. E. & Kelly, C. N. Effect of formation conditions on biochars: Compositional and structural properties of cellulose, lignin, and pine biochars. Biomass Bioenerg. 46, 693–701 (2012).
CAS  Article  Google Scholar 

36.
Kastner, J. R., Miller, J. & Das, K. C. Pyrolysis conditions and ozone oxidation effects on ammonia adsorption in biomass generated chars. J. Hazard. Mater. 164, 1420–1427 (2009).
CAS  Article  Google Scholar 

37.
Zhao, L., Cao, X., Mašek, O. & Zimmerman, A. Heterogeneity of biochar properties as a function of feedstock sources and production temperatures. J. Hazard. Mater. 256–257, 1–9 (2013).
PubMed  PubMed Central  Google Scholar 

38.
Lehmann, J. A handful of carbon. Nature 447, 143–144 (2007).
ADS  CAS  Article  Google Scholar 

39.
de la Rosa, J. M., Rosado, M., Paneque, M., Miller, A. Z. & Knicker, H. Effects of aging under field conditions on biochar structure and composition: Implications for biochar stability in soils. Sci. Total Environ. 613–614, 969–976 (2018).
Article  CAS  Google Scholar 

40.
Huang, Z. et al. Effect of aging on surface chemistry of rice husk-derived biochar. Environ. Prog. Sustain. Energy. 37, 410–417 (2017).
Article  CAS  Google Scholar 

41.
Mia, S., Dijkstra, F. A. & Singh, B. Aging induced changes in biochar’s functionality and adsorption behavior for phosphate and ammonium. Environ. Sci. Technol. 51, 8359–8367 (2017).
ADS  CAS  Article  Google Scholar 

42.
Zhou, B. et al. Impact of hydrochar on rice paddy CH4 and N2O emissions: A comparative study with pyrochar. Chemosphere 204, 474–482 (2018).
ADS  CAS  Article  Google Scholar 

43.
Sun, X., Zhong, T., Zhang, L., Zhang, K. & Wu, W. Reducing ammonia volatilization from paddy field with rice straw derived biochar. Sci. Total Environ. 660, 512–518 (2019).
ADS  CAS  Article  Google Scholar  More

Food sustainability indicators
The indicators of the five dimensions of food sustainability that were collectively defined and assessed in the six food systems are presented in Table 2 and in the Supplementary Data (sheets 1–9). Relevant across contexts, the indicators represent a consensual output of the research process with scientists from the Global North and South and non-academic actors related to the different food systems (see Methods). The indicators cover different activities, from production to consumption, and some are transversal, i.e. occurring along the value chain.
How the six food systems scored
Food systems B3 (Agroecological food system) and K3 (Local food system) had the highest overall sustainability scores. In addition, these scores were more equally distributed across the five dimensions than in the other food systems. The greatest contributor to these high scores was environmental performance: both food systems demonstrated a high capacity to provide agroecosystem services (e.g. through crop diversity or combining livestock with trees29,33); low external inputs and recycling of organic materials; a low carbon footprint; and perceived positive health impacts by producers, workers and consumers. The food system that scored highest (4.0) in environmental performance is the Domestic–indigenous food system (B2). However, it obtained the lowest scores in poverty and inequality (1.6, with particularly low ratings for incomes, livelihood capitals and social protection), pulling down its overall score.
Figures 1 and 2 display the aggregated qualitative and quantitative research results on a five-point Likert scale from 0 (very low) to 4 (very high). The area covered by one food system reflects its overall sustainability, while the axes reflect the five dimensions. The median is calculated as an average value for one dimension from all its indicators; for each food system it represents strengths (comparatively high scores) and weaknesses (comparatively low scores) of food sustainability of the six assessed food systems.
Figure 1

Overall food sustainability scores and median scores of five dimensions for three food systems in Kenya, rated from 0 (very low), 1 (low), 2 (medium), 3 (high) to 4 (very high). For detailed results, see Supplementary Data.

Full size image

Figure 2

Overall food sustainability scores and median scores of five dimensions for three food systems in Bolivia, from 0 (very low), 1 (low), 2 (medium), 3 (high) to 4 (very high). For detailed results, see Supplementary Data.

Full size image

The lowest overall sustainability scores were obtained by the Agro-industrial food systems, B1 (scoring 1.6) and K1 (scoring 1.8). This was mainly due to their poor environmental performance on pesticide and resource use. Of the pesticides documented during this study, 65% in Bolivia and 67% in Kenya contained substances considered “highly hazardous” jointly by the FAO and WHO36. Additionally, resource use along the value chain was high, with examples including water, packaging material, electricity and diesel, and, in Kenya, aviation turbine oil30,37. Lowest-scoring B1 demonstrated a low diversity of crops and breeds, high greenhouse gas emissions and perceived negative health impacts. Right to food was particularly low in B1 due to low quality and accessibility of land and water resources for the local population, low food diversity and access to seeds, low access for women to land and finance, and a lack of participation in decision-making. In second-lowest scoring K1, water use was around 100 times higher than in K3, pesticide use seven times higher, and the carbon footprint of exported vegetables 67 times higher than for vegetables consumed in K330.
Food security of local households in the study areas was highest in the Agroecological food system (B3), with better scores than the other food systems for access to land and water, contribution to local consumption, accessibility of food, and capacity to provide what is considered to constitute a “good diet”. In general, household food security was high in the study area in Bolivia, and low to medium in the study area in Kenya. Food security was lowest in the Agro-industrial system in Kenya (K1). This is because K1 exports almost all the food it produces and does not engage in processing or storage activities, implying low accessibility to, and consumption of, the produced food locally. Households involved in K1 through labour had medium food security and a low perception of the food system’s capacity to provide a “good diet”.
Contrary to expectations, the Agro-industrial food systems obtained medium (B1) and above-medium (K1) resilience scores. Key factors were a high or very high level of self-organization in interest groups, knowledge on threats and opportunities, and functioning feedback mechanisms between system components, such as supportive policies that translated into subsidies, relief payments and reduced tax rates24. This social dimension of resilience somewhat mitigated the low scores that B1 and K1 obtained for agroecosystem resilience and their high dependence on external inputs and monocultures (which, in turn, rendered them vulnerable to e.g. climate impacts or price fluctuations).
The weakest dimensions across food systems
The weakest dimension was right to food. K1 and B1 both scored particularly low in this dimension due to high land concentration (e.g. average land plot size was 90 ha in K1, compared to 2 ha in K324) and a lack of food diversity, supply of nutritional needs, and local food traditions. All food systems obtained low scores for women’s access to land and credit (in Kenya, only 5% and in Bolivia 17% of landowners are women38). K3 obtained slightly higher scores, as more women had access to land (although this did not mean they held the property deeds) and because of the prevalence of women’s groups that operated a system of microcredits.
The second-weakest dimension was poverty and inequality. This was due to low farming incomes and high income inequality (e.g. salaries for selling agricultural inputs in B1 were 220% higher than for the other activities in this food system24). Salaries for workers (e.g. farm workers in Kenya39) were around the minimum wage, and workers throughout the value chain were excluded from social protection. Nevertheless, the Agro-industrial food systems obtained a high (K1) and a medium score (B1) for the reduction of poverty and inequality, due to high scores for physical capital (infrastructure, fulfilment of basic needs, transport and storage facilities, livestock) and human capital (education, experience, health), and relatively low household expenditure on food.
Contributions of food system activities to sustainability
To understand the contribution of different food system activities to the overall sustainability scores, the indicators for each food system are grouped according to activity: production, processing and storage, retail and trade, consumption, and transversal (across activities, e.g. carbon footprint of a food product). Figure 3 shows the sum of the medians according to activity, and Fig. 4 shows the range of scores for each activity in each food system.
Figure 3

Median food system activity score of food sustainability. “Transversal” means across all food system activities. The maximum score for each food system activity is 4 (or “very high” on the Likert scale), and the overall maximum score is 20.

Full size image

Figure 4

Distribution of sustainability scores for each food system according to food system activity: production, processing/storage, retail/trade, consumption and transversal indicators.

Full size image

In the food systems with a comparably high overall sustainability score (B3, K3), all activities obtained relatively high scores (e.g. consumption in B3: locally produced food, provision of food to food system actors, a perceived “good diet”, contribution to food diversity, information and participation). The “transversal” category recorded similar scores across food systems. It comprised household food security, livelihood assets, material and energy use along the value chain, and resilience indicators (e.g. organization in interest groups, also along the value chain). The food system with the lowest cumulative score, K1, scored 0 in processing/storage and retail/trade, and it obtained low scores for production (due to low incomes), access to productive resources, environmental performance, and consumption (due to low contributions to the local food system and its diversity). Transversal scored higher than the other activities in K1, mainly due to the positive social resilience scores mentioned above. Figure 4 shows the per-activity contribution to the overall food sustainability rating for each food system.
Most food system activities (especially production, consumption and transversal) had a high variability of scores, ranging from 0–3 or even from 0–4 (minimum to maximum value). In B3, every activity obtained a comparably high score, although all but retail and trade were still very variable. Processing and storage (capacity in the food system to provide both processing and storage) was medium to high in B2, but storage was low in K3 and B3 (weakening overall food security) and K1 (freshly sold perishable produce). Retail and trade (affordable food prices, above-medium retail employee wages) contributed strongly to overall food sustainability in B3 and K3, at a medium level to B1 and K2, and little to B2. Consumption obtained a medium or above-medium score, which means that it played an important role in overall sustainability (e.g. in the form of food diversity in K3). An exception was K1, where consumption took place so far away that most of the related indicators obtained low scores for the food system context under study. Scores obtained for the “transversal” category also varied highly, but augmented overall food sustainability mainly through resilience (K1, B1) and environmental performance indicators (K2, K3, B3, K3).
Most decisive indicators for food sustainability
To identify general trends, we further analysed the importance of individual indicators for overall food sustainability across all six food systems (Fig. 5).
Figure 5

Frequency of difference from the median (to the left of 0: frequency with which the indicators across all six food systems scored worse than the median; to the right of 0: frequency with which they scored better than the median).

Full size image

Resilience indicators often had a strongly positive influence on overall sustainability, especially the knowledge of threats and opportunities indicator, with above-median scores in five of the six food systems, and the indicators on functioning feedback mechanisms, interest groups and shared vision, which achieved above-median scores in four of the food systems. A notable exception was diversity of crops and breeds, a resilience indicator which scored lower than the overall median in five food systems. Several indicators from the food security dimension scored better than the median in four food systems: ability to provide food to food system actors, capacity to process food, access to land, access to water, and household food security.
In addition, environmental performance indicators were often high (e.g. use of energy, soil quality, use of materials, water footprint, Agroecosystem Service Capacity Index, carbon footprint), but mainly for the more local and diversified food systems. Exceptions were formed by the water footprint and use of materials along the value chain or food system stages, which were low also in B1 (as calculated up to the first consumption stage, e.g. use of soybeans for feed in meat and dairy production).
Low-scoring dimensions—those that pulled down the overall food sustainability score, i.e. poverty and inequality, and right to food—included indicators that most frequently scored lower than the median. These were related to gender, dwindling agrobiodiversity and food diversity, and precarious work conditions at the production level. Of these, women’s access to credit and diversity of crops and breeds scored five times below the median. The second-worst indicators (four times below the median) with no positive score were social protection and local food traditions, and the third-worst were the proportion of women with land rights, remedies for violations of the right to food, and liveable wage.
A principal component analysis (PCA) providing information on which combinations of indicators are most decisive for overall food sustainability in our case studies confirmed the trend shown in Fig. 5. Four principal components retained based on their eigenvalues explained 99% of variance (Supplementary Table S1). By retaining indicators with component loadings  > 0.45, the first principal component was most influenced by human capital, social protection, remedies for violations of the right to food, local food traditions, access to information, landscape heterogeneity, water quality for domestic consumption, and women’s access to credit. Most of these indicators belong to the right to food and poverty inequality dimensions and are related to diversity and quality of human and natural resources that households, and especially women, have access to. The second principal component was most influenced by the capacity to process food, accessibility of water for domestic consumption, farmer incomes, ability to provide food to food system actors, use of materials, use of energy, and the capacity to cover nutritional needs, and was thus mainly linked to environmental performance and food security. The third principal component was dominated by resilience indicators: interest groups, knowledge of threats and opportunities, decent and safe working conditions, use of energy, shared vision, and ecological self-regulation. The fourth principal component was mostly influenced by access to water for domestic consumption and for irrigation, wages in retail, household food security, the proportion of women with land rights, and reflective and shared learning, and was thus strongly related to access to resources and incomes, particularly for women.
From the two analyses (frequency of positive/negative scoring of indicators, and PCA), we can identify the indicators with the greatest influence across the food systems under study. Six of these contributed positively, meaning that they were in a rather good state in several of the food systems. Four of the six indicators were from the food security dimension (capacity of the food system to process food, ability of the food system to provide food to food system actors, household food security, access to water) and two were from the environmental performance dimension (use of materials and use of energy). We identify six indicators, all but diversity of crops and breeds from the right to food dimension, which had a strongly negative influence (women’s access to credit, social protection, local food traditions, women’s land rights, and remedies for violations of the right to food). This means that these indicators were in an undesirable state in most of the food systems. More

1.
Gandhi, S. & Jones, T. G. Identifying mangrove deforestation hotspots in South Asia, Southeast Asia and Asia-Pacific. Remote Sens. 11, 728 (2019).
ADS  Article  Google Scholar 
2.
Hamdan, O., Khali-Aziz, H., Shamsudin, I. & Raja-Barizan, R.S. Status of Mangroves in Peninsular Malaysia. 153 (Forest Research Institute Malaysia, 2012).

3.
Taillardat, P., Friess, D. A. & Lupascu, M. Mangrove blue carbon strategies for climate change mitigation are most effective at the national scale. Biol. Lett. 14, 20180251 (2018).
Article  CAS  Google Scholar 

4.
Richards, D. R. & Friess, D. A. Rates and drivers of mangrove deforestation in Southeast Asia, 2000–2012. Proc. Natl. Acad. Sci. USA 113, 344–349 (2016).
ADS  CAS  Article  Google Scholar 

5.
Friess, D. A. et al. Mangroves give cause for conservation optimism, for now. Curr. Biol. 30, R153–R154 (2020).
CAS  Article  Google Scholar 

6.
Polidoro, B. A. et al. The loss of species: mangrove extinction risk and geographic areas of global concern. PLoS ONE 5, e10095 (2010).
ADS  Article  CAS  Google Scholar 

7.
Matesanz, S., Rubio-Teso, M. L., García-Fernández, A. & Escudero, A. Habitat fragmentation differentially affects genetic variation, phenotypic plasticity and survival in populations of a gypsum endemic. Front. Plant Sci. 8, 843 (2017).
Article  Google Scholar 

8.
Furches, M. S., Small, R. L. & Furches, A. Genetic diversity in three endangered pitcher plant species (Sarracenia; Sarraceniaceae) is lower than widespread congeners. Am. J. Bot. 100, 2092–2101 (2013).
Article  Google Scholar 

9.
Yan, Y. B., Duke, N. C. & Sun, M. Comparative analysis of the pattern of population genetic diversity in three Indo-West Pacific Rhizophora mangrove species. Front. Plant Sci. 7, 1434 (2016).
PubMed  Google Scholar 

10.
Wan-Ismail, W. N., Wan-Ahmad, W. J., Salam, M. R. & Latiff, A. Structural and floristic pattern in a disturbed mangrove tropical swamp forest: a case study from the Langkawi UNESCO Global Geopark Forest, Peninsular Malaysia. Sains Malays. 47, 861–869 (2018).
Article  Google Scholar 

11.
Setyawan, A.D., Ulumuddin, Y.I. & Ragavan, P. Mangrove hybrid of Rhizophora and its parentals species in Indo-Malayan region. Nusantara Biosci. 6 (2014).

12.
Lahjie, A.M., Nouval, B., Lahjie, A.A., Ruslim, Y. & Kristiningrum, R. Economic valuation from direct use of mangrove forest restoration in Balikpapan Bay, East Kalimantan, Indonesia. F1000Res. 8 (2019).

13.
Omar, H., Misman, M.A. & Musa, S. GIS and remote sensing for mangroves mapping and monitoring. Geographic Information Systems and Science. IntechOpen https://www.intechopen.com/books/geographic-information-systems-and-science/gis-and-remote-sensing-for-mangroves-mapping-and-monitoring (2019).

14.
Takayama, K., Tamura, M., Tateishi, Y., Webb, E. L. & Kajita, T. Strong genetic structure over the American continents and transoceanic dispersal in the mangrove genus Rhizophora (Rhizophoraceae) revealed by broad-scale nuclear and chloroplast DNA analysis. Am. J. Bot. 100, 1191–1201 (2013).
CAS  Article  Google Scholar 

15.
Ng, W. L. et al. Closely related and sympatric but not all the same: genetic variation of Indo-West Pacific Rhizophora mangroves across the Malay Peninsula. Conserv. Genet. 16, 137–150 (2015).
Article  Google Scholar 

16.
Yahya, A. F. et al. Genetic variation and population genetic structure of Rhizophora apiculata (Rhizophoraceae) in the greater Sunda Islands, Indonesia using microsatellite markers. J. Plant Res. 127, 287–297 (2014).
CAS  Article  Google Scholar 

17.
Chen, Y. et al. Applications of multiple nuclear genes to the molecular phylogeny, population genetics and hybrid identification in the mangrove genus Rhizophora. PLoS ONE. 10 (2015).

18.
Guo, Z. et al. Genetic discontinuities in a dominant mangrove Rhizophora apiculata (Rhizophoraceae) in the Indo-Malesian region. J. Biogeogr. 43, 1856–1868 (2016).
Article  Google Scholar 

19.
Cheng, A. et al. Molecular marker technology for genetic improvement of underutilised crops. In Crop improvement (eds Abdullah, S. et al.) 47–70 (Springer, Cham, 2017).
Google Scholar 

20.
Ali, A. et al. Genetic diversity and population structure analysis of Saccharum and Erianthus genera using microsatellite (SSR) markers. Sci. Rep. 9, 1–10 (2019).
Article  CAS  Google Scholar 

21.
Shinmura, Y. et al. Isolation and characterization of 14 microsatellite markers for Rhizophora mucronata (Rhizophoraceae) and their potential use in range-wide population studies. Conserv. Genet. Resour. 4, 951–954 (2012).
Article  Google Scholar 

22.
Xu, S. et al. The origin, diversification and adaptation of a major mangrove clade (Rhizophoraceae) revealed by whole-genome sequencing. Natl. Sci. Rev. 4, 721–734 (2017).
CAS  Article  Google Scholar 

23.
Evanno, G., Regnaut, S. & Goudet, J. Detecting the number of clusters of individuals using the software STRUCTURE: a simulation study. Mol. Ecol. 14, 2611–2620 (2005).
CAS  Article  Google Scholar 

24.
Nei, M., Tajima, F. & Tateno, Y. Accuracy of estimated phylogenetic trees from molecular data. J. Mol. Evol. 19, 153–170 (1983).
ADS  CAS  Article  Google Scholar 

25.
Maguire, T.L., Edwards, K.J., Saenger, P. & Henry, R. Characterisation and analysis of microsatellite loci in a mangrove species, Avicennia marina (Forsk.) Vierh. (Avicenniaceae). Theor. Appl. Genet. 101, 279–285 (2000).

26.
Torre, S. et al. RNA-seq analysis of Quercus pubescens leaves: de novo transcriptome assembly, annotation and functional markers development. PLoS ONE 9, e112487 (2014).
ADS  Article  CAS  Google Scholar 

27.
Ye, Y. et al. Characterization, validation, and cross-species transferability of newly developed EST-SSR markers and their application for genetic evaluation in crape myrtle (Lagerstroemia spp). Mol. Breed. 39, 26 (2019).
Article  CAS  Google Scholar 

28.
Nei, M. Molecular Evolutionary Genetics (Columbia University Press, London, 1987).
Google Scholar 

29.
Wee, A. K. et al. Vicariance and oceanic barriers drive contemporary genetic structure of widespread mangrove species Sonneratia alba, J. Sm in the Indo-West Pacific. Forests 8, 483 (2017).
Article  Google Scholar 

30.
Ellstrand, N. C. & Elam, D. R. Population genetic consequences of small population size: implications for plant conservation. Annu. Rev. Ecol. Evol. Syst. 24, 217–242 (1993).
Article  Google Scholar 

31.
Feder, J. L., Gejji, R., Yeaman, S. & Nosil, P. Establishment of new mutations under divergence and genome hitchhiking. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 367, 461–474 (2012).
Article  Google Scholar 

32.
Annuar, A. S. & Latip, N. A. Mangrove contributions towards environmental conservation and tourism in Balik Pulau. Adv. Conserv. Sci. Technol. 1, 1–7 (2020).
Google Scholar 

33.
Wee, A.K. et al. Oceanic currents, not land masses, maintain the genetic structure of the mangrove Rhizophora mucronata Lam. (Rhizophoraceae) in Southeast Asia. J. Biogeogr. 41, 954–964 (2014).

34.
Ismail, M. H., Zaki, P. H. & Hamed, A. A. Wood density and carbon estimates of mangrove species in Kuala Sepetang, Perak, Malaysia. Malays. For. 78, 115–124 (2015).
Google Scholar 

35.
Vitorino, C. A., Nogueira, F., Souza, I. L., Araripe, J. & Venere, P. C. Low genetic diversity and structuring of the Arapaima (Osteoglossiformes, Arapaimidae) population of the Araguaia-Tocantins basin. Front. Genet. 8, 159 (2017).
Article  Google Scholar 

36.
Wright, S. The genetical structure of populations. Ann. Eugen. 15, 323–354 (1951).
MathSciNet  CAS  MATH  Article  Google Scholar 

37.
Slatkin, M. A measure of population subdivision based on microsatellite allele frequencies. Genetics 139, 457–462 (1995).
CAS  PubMed  Google Scholar 

38.
Goodman, S. J. RST calc: a collection of computer programs for calculating estimates of genetic differentiation from microsatellite data and determining their significance. Mol. Ecol. 6, 881–885 (1997).
CAS  Article  Google Scholar 

39.
Moulin, N. L., Wyttenbach, A., Brüunner, H., Goudet, J. & Hausser, J. Study of gene flow through a hybrid zone in the common shrew (Sorex araneus) using microsatellites. Hereditas. 125, 159–168 (1996).
Article  Google Scholar 

40.
Ge, X. J. & Sun, M. Population genetic structure of Ceriops tagal (Rhizophoraceae) in Thailand and China. Wetl. Ecol. Manag. 9, 213–219 (2001).
Article  Google Scholar 

41.
Dodd, R.S., Afzal-Rafii, Z., Kashani, N. & Budrick, J. Land barriers and open oceans: effects on gene diversity and population structure in Avicennia germinans L. (Avicenniaceae). Mol. Ecol. 11, 1327–1338 (2002).

42.
Rizal, S. et al. General circulation in the Malacca strait and Andaman Sea: a numerical model study. Am. J. Environ. Sci. 8, 479–488 (2012).
Article  Google Scholar 

43.
Nathan, R. et al. Mechanisms of long-distance seed dispersal. Trends Ecol. Evol. 23, 638–647 (2008).
Article  Google Scholar 

44.
Drexler, J.Z. Maximum longevities of Rhizophora apiculata and R. mucronata propagules. Pac. Sci. 55, 17–22 (2001).

45.
Li, J. et al. Pronounced genetic differentiation and recent secondary contact in the mangrove tree Lumnitzera racemosa revealed by population genomic analyses. Sci. Rep. 6, 29486 (2016).
ADS  CAS  Article  Google Scholar 

46.
Murray, M. G. & Thompson, W. F. Rapid isolation of high molecular weight plant DNA. Nucl. Acids Res. 8, 4321–4326 (1980).
CAS  Article  Google Scholar 

47.
Andrews, S. FastQC: a quality control tool for high throughput sequence data. Babraham Bioinformatics https://www.bioinformatics.babraham.ac.uk/projects/fastqc (2010).

48.
Bolger, A. M., Lohse, M. & Usadel, B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinform. 30, 2114–2120 (2014).
CAS  Article  Google Scholar 

49.
Grabherr, M. G. et al. Trinity: reconstructing a full-length transcriptome without a genome from RNA-Seq data. Nat. Biotechnol. 29, 644 (2011).
CAS  Article  Google Scholar 

50.
Varshney, R. K., Thiel, T., Stein, N., Langridge, P. & Graner, A. In silico analysis on frequency and distribution of microsatellites in ESTs of some cereal species. Cell Mol. Biol. Lett. 7, 537–546 (2002).
CAS  PubMed  Google Scholar 

51.
Rozen, S. & Skaletsky, H. Primer3 on the WWW for general users and for biologist programmers in Bioinformatics methods and protocols. 365–386 (Humana Press, 2000).

52.
Van-Oosterhout, C., Hutchinson, W. F., Wills, D. P. & Shipley, P. MICRO-CHECKER: software for identifying and correcting genotyping errors in microsatellite data. Mol. Ecol. Notes. 4, 535–538 (2004).
CAS  Article  Google Scholar 

53.
Lewis, P.O. & Zaykin, D. Genetic Data Analysis (GDA) version 1.1: a computer program for the analysis of allelic data. UConn https://phylogeny.uconn.edu/software/ (2002).

54.
Rice, W. R. Analyzing tables of statistical tests. Evol. 43, 223–225 (1989).
Article  Google Scholar 

55.
Park, S.D.E. Trypanotolerance in West African cattle and the population genetic effects of selection. Ph. D (University of Dublin, 2001).

56.
Goudet, J. FSTAT version 2.9.3.2: a program to estimate and test gene diversities and fixation indices. Unil https://www2.unil.ch/popgen/softwares/fstat.htm (2002).

57.
Nei, M., Tajima, F. & Tateno, Y. Accuracy of estimated phylogenetic trees from molecular data. J. Mol. Evol. 19, 153–170 (1983).

58.
Liu, K. & Muse, S. V. PowerMarker: an integrated analysis environment for genetic marker analysis. Bioinform. 21, 2128–2129 (2005).
CAS  Article  Google Scholar 

59.
Nei, M. F-statistics and analysis of gene diversity in subdivided populations. Ann. Hum. Genet. 41, 225–233 (1977).
CAS  MATH  Article  Google Scholar 

60.
Peakall, R. & Smouse, P. E. GenAlEx 6.5: genetic analysis in Excel. Population genetic software for teaching and research-an update. Bioinform. 28, 2537–2539 (2012).
CAS  Article  Google Scholar 

61.
Pritchard, J. K., Stephens, M. & Donnelly, P. Inference of population structure using multilocus genotype data. Genetics 155, 945–959 (2000).
CAS  PubMed  Google Scholar 

62.
Li, Y. L. & Liu, J. X. StructureSelector: a web-based software to select and visualize the optimal number of clusters using multiple methods. Mol. Ecol. Resour. 18, 176–177 (2018).
Article  Google Scholar 

63.
Goudet, J. PCAGEN version 1.2: a program to perform a principal component analysis (PCA) on genetic data. Unil https://www2.unil.ch/popgen/softwares/pcagen.htm (1999).

64.
Tamura, K. et al. MEGA5: molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Mol. Biol. Evol. 28, 2731–2739 (2011).
CAS  Article  Google Scholar  More

1.
Arndt S, Jørgensen BB, LaRowe DE, Middelburg J, Pancost R, Regnier P. Quantifying the degradation of organic matter in marine sediments: a review and synthesis. Earth Sci Rev. 2013;123:53–86.
CAS  Article  Google Scholar 
2.
Froelich PN, Klinkhammer GP, Bender ML, Luedtke NA, Heath GR, Cullen D, et al. Early oxidation of organic matter in pelagic sediments of the eastern equatorial Atlantic: suboxic diagenesis. Geochim Cosmochim Acta. 1979;43:1075–90.
CAS  Article  Google Scholar 

3.
Calvert S. Oceanographic controls on the accumulation of organic matter in marine sediments. Geol Soc Spec Publ. 1987;26:137–51.
Article  Google Scholar 

4.
De Leeuw J, Largeau C. A review of macromolecular organic compounds that comprise living organisms and their role in kerogen, coal, and petroleum formation. Org Geochem. 1993;11:23–72.

5.
Mackenzie FT, Lerman A, Andersson AJ. Past and present of sediment and carbon biogeochemical cycling models. Biogeosciences. 2004;1:11–32.
CAS  Article  Google Scholar 

6.
Oni OE, Miyatake T, Kasten S, Richter-Heitmann T, Fischer D, Wagenknecht L, et al. Distinct microbial populations are tightly linked to the profile of dissolved iron in the methanic sediments of the Helgoland Mud Area, North Sea. Front Microbiol. 2015;6:365.
PubMed Central  PubMed  Google Scholar 

7.
Egger M, Hagens M, Sapart CJ, Dijkstra N, van Helmond NA, Mogollón JM, et al. Iron oxide reduction in methane-rich deep Baltic Sea sediments. Geochim Cosmochim Acta. 2017;207:256–76.
CAS  Article  Google Scholar 

8.
Riedinger N, Pfeifer K, Kasten S, Garming JFL, Vogt C, Hensen C. Diagenetic alteration of magnetic signals by anaerobic oxidation of methane related to a change in sedimentation rate. Geochim Cosmoch Acta. 2005;69:4117–26.
CAS  Article  Google Scholar 

9.
Riedinger N, Formolo MJ, Lyons TW, Henkel S, Beck A, Kasten S. An inorganic geochemical argument for coupled anaerobic oxidation of methane and iron reduction in marine sediments. Geobiology. 2014;12:172–81.
CAS  Article  Google Scholar 

10.
März C, Hoffmann J, Bleil U, De Lange G, Kasten S. Diagenetic changes of magnetic and geochemical signals by anaerobic methane oxidation in sediments of the Zambezi deep-sea fan (SW Indian Ocean). Mar Geol. 2008;255:118–30.
Article  CAS  Google Scholar 

11.
Hensen C, Zabel M, Pfeifer K, Schwenk T, Kasten S, Riedinger N, et al. Control of sulfate pore-water profiles by sedimentary events and the significance of anaerobic oxidation of methane for the burial of sulfur in marine sediments. Geochim Cosmochim Acta. 2003;67:2631–47.
CAS  Article  Google Scholar 

12.
Flood RD, Piper DJW, Klaus A, Party SS. Initial Reports. Proc. Ocean Drill. Progam. 1995;155. https://doi.org/10.2973/odp.proc.ir.155.1995.

13.
Kasten S, Freudenthal T, Gingele FX, Schulz HD. Simultaneous formation of iron-rich layers at different redox boundaries in sediments of the Amazon deep-sea fan. Geochim Cosmochim Acta. 1998;62:2253–64.
CAS  Article  Google Scholar 

14.
Meyers SR. Production and preservation of organic matter: the significance of iron. Paleoceanography. 2007;22:PA4211.

15.
Barber A, Brandes J, Leri A, Lalonde K, Balind K, Wirick S, et al. Preservation of organic matter in marine sediments by inner-sphere interactions with reactive iron. Sci Rep. 2017;7:1–10.
CAS  Article  Google Scholar 

16.
Lalonde K, Mucci A, Ouellet A, Gélinas Y. Preservation of organic matter in sediments promoted by iron. Nature. 2012;483:198–200.
CAS  Article  Google Scholar 

17.
Middelburg JJ. A simple rate model for organic matter decomposition in marine sediments. Geochim Cosmochim Acta. 1989;53:1577–81.
CAS  Article  Google Scholar 

18.
Biddle JF, Lipp JS, Lever MA, Lloyd KG, Sørensen KB, Anderson R, et al. Heterotrophic Archaea dominate sedimentary subsurface ecosystems off Peru. Proc Natl Acad Sci USA. 2006;103:3846–51.
CAS  Article  Google Scholar 

19.
Aromokeye DA, Kulkarni AC, Elvert M, Wegener G, Henkel S, Coffinet S, et al. Rates and microbial players of iron-driven anaerobic oxidation of methane in methanic marine sediments. Front Microbiol. 2020;10:3041.
PubMed Central  Article  PubMed  Google Scholar 

20.
Lovley DR, Phillips EJ. Organic matter mineralization with reduction of ferric iron in anaerobic sediments. Appl Environ Microbiol. 1986;51:683–9.
CAS  PubMed Central  Article  PubMed  Google Scholar 

21.
Lovley DR, Coates JD, Blunt-Harris EL, Phillips EJ, Woodward JC. Humic substances as electron acceptors for microbial respiration. Nature. 1996;382:445–8.
CAS  Article  Google Scholar 

22.
Lovley D. Dissimilatory Fe (III)-and Mn (IV)-reducing prokaryotes, In: Dworkin M, Falkow S, Rosenberg E, Schleifer K-H, Stackebrandt E (eds) The Prokaryotes. Springer: Berlin Heidelberg; 2006, Vol. 2, p. 635–58.

23.
Lovley DR, Phillips EJ. Novel mode of microbial energy metabolism: organic carbon oxidation coupled to dissimilatory reduction of iron or manganese. Appl Environ Microbiol. 1988;54:1472–80.
CAS  PubMed Central  Article  PubMed  Google Scholar 

24.
Kato S, Hashimoto K, Watanabe K. Methanogenesis facilitated by electric syntrophy via (semi) conductive iron‐oxide minerals. Environ Microbiol. 2012;14:1646–54.
CAS  Article  Google Scholar 

25.
Jiang S, Park S, Yoon Y, Lee J-H, Wu W-M, Phuoc Dan N, et al. Methanogenesis facilitated by geobiochemical iron cycle in a novel syntrophic methanogenic microbial community. Environ Sci Technol. 2013;47:10078–84.
CAS  Article  Google Scholar 

26.
Aromokeye DA, Richter-Heitmann T, Oni O, Emmanuel, Kulkarni A, Yin X, et al. Temperature controls crystalline iron oxide utilization by microbial communities in methanic ferruginous marine sediment incubations. Front Microbiol. 2018;9:2574.
PubMed Central  Article  PubMed  Google Scholar 

27.
Zhuang L, Tang J, Wang Y, Hu M, Zhou S. Conductive iron oxide minerals accelerate syntrophic cooperation in methanogenic benzoate degradation. J Hazard Mater. 2015;293:37–45.
CAS  Article  Google Scholar 

28.
Hebbeln D, Scheurle C, Lamy F. Depositional history of the Helgoland Mud Area, German Bight, North Sea. Geo Mar Lett. 2003;23:81–90.
Article  Google Scholar 

29.
Oni OE, Schmidt F, Miyatake T, Kasten S, Witt M, Hinrichs K-U, et al. Microbial communities and organic matter composition in surface and subsurface sediments of the Helgoland Mud Area, North Sea. Front Microbiol. 2015;6:1290.
PubMed Central  PubMed  Google Scholar 

30.
Gan S, Schmidt F, Heuer VB, Goldhammer T, Witt M, Hinrichs K-U. Impacts of redox conditions on dissolved organic matter (DOM) quality in marine sediments off the River Rhône, Western Mediterranean Sea. Geochim Cosmochim Acta. 2020;276:151–69.
CAS  Article  Google Scholar 

31.
Carmona M, Zamarro MT, Blázquez B, Durante-Rodríguez G, Juárez JF, Valderrama JA, et al. Anaerobic catabolism of aromatic compounds: a genetic and genomic view. Microbiol Mol Biol Rev. 2009;73:71–133.
CAS  PubMed Central  Article  PubMed  Google Scholar 

32.
Fuchs G, Boll M, Heider J. Microbial degradation of aromatic compounds—from one strategy to four. Nat Rev Microbiol. 2011;9:803–16.
CAS  Article  Google Scholar 

33.
Gibson J, Harwood SC. Metabolic diversity in aromatic compound utilization by anaerobic microbes. Annu Rev Microbiol. 2002;56:345–69.
CAS  Article  Google Scholar 

34.
Hopkins BT, McInerney MJ, Warikoo V. Evidence for anaerobic syntrophic benzoate degradation threshold and isolation of the syntrophic benzoate degrader. Appl Environ Microbiol. 1995;61:526.
CAS  PubMed Central  Article  PubMed  Google Scholar 

35.
Schink B. Energetics of syntrophic cooperation in methanogenic degradation. Microbiol Mol Biol Rev. 1997;61:262–80.
CAS  PubMed Central  Article  PubMed  Google Scholar 

36.
Schöcke L, Schink B. Energetics of methanogenic benzoate degradation by Syntrophus gentianae in syntrophic coculture. Microbiology. 1997;143:2345–51.
Article  Google Scholar 

37.
Widdel F, Kohring G-W, Mayer F. Studies on dissimilatory sulfate-reducing bacteria that decompose fatty acids III. Characterization of the filamentous gliding Desulfonema limicola gen. nov. sp. nov., and Desulfonema magnum sp. nov. Arch Microbiol. 1983;134:286–94.
CAS  Article  Google Scholar 

38.
Widdel F. Anaerober Abbau von Fettsäuren und Benzoesäure durch neu isolierte Arten sulfat-reduzierender Bakterien [PhD Thesis]. Göttingen, Germany: Georg-August-Universität zu Göttingen; 1980.

39.
Widdel F, Pfennig N. Studies on dissimilatory sulfate-reducing bacteria that decompose fatty acids. Arch Microbiol. 1981;129:395–400.
CAS  Article  Google Scholar 

40.
Viollier E, Inglett P, Hunter K, Roychoudhury A, Van, Cappellen P. The ferrozine method revisited: Fe(II)/Fe(III) determination in natural waters. Appl Geochem. 2000;15:785–90.
CAS  Article  Google Scholar 

41.
Heuer VB, Pohlman JW, Torres ME, Elvert M, Hinrichs K-U. The stable carbon isotope biogeochemistry of acetate and other dissolved carbon species in deep subseafloor sediments at the northern Cascadia Margin. Geochim Cosmochim Acta. 2009;73:3323–36.
CAS  Article  Google Scholar 

42.
Lin Y-S, Heuer VB, Goldhammer T, Kellermann MY, Zabel M, Hinrichs K-U. Towards constraining H2 concentration in subseafloor sediment: a proposal for combined analysis by two distinct approaches. Geochim Cosmochim Acta. 2012;77:186–201.
CAS  Article  Google Scholar 

43.
Lueders T, Manefield M, Friedrich MW. Enhanced sensitivity of DNA‐and rRNA‐based stable isotope probing by fractionation and quantitative analysis of isopycnic centrifugation gradients. Environ Microbiol. 2004;6:73–8.
CAS  Article  Google Scholar 

44.
Amann R, Fuchs BM, Behrens S. The identification of microorganisms by fluorescence in situ hybridisation. Curr Opin Biotechnol. 2001;12:231–6.
CAS  Article  Google Scholar 

45.
Poulton SW, Krom MD, Raiswell R. A revised scheme for the reactivity of iron (oxyhydr)oxide minerals towards dissolved sulfide. Geochim Cosmochim Acta. 2004;68:3703–15.
CAS  Article  Google Scholar 

46.
Herndon EM, Yang Z, Bargar J, Janot N, Regier TZ, Graham DE, et al. Geochemical drivers of organic matter decomposition in arctic tundra soils. Biogeochemistry. 2015;126:397–414.
CAS  Article  Google Scholar 

47.
Yang Z, Wullschleger SD, Liang L, Graham DE, Gu B. Effects of warming on the degradation and production of low-molecular-weight labile organic carbon in an Arctic tundra soil. Soil Biol Biochem. 2016;95:202–11.
CAS  Article  Google Scholar 

48.
Yang Z, Shi X, Wang C, Wang L, Guo R. Magnetite nanoparticles facilitate methane production from ethanol via acting as electron acceptors. Sci Rep. 2015;5;16118. https://doi.org/10.1038/srep16118.

49.
McInerney MJ, Struchtemeyer CG, Sieber J, Mouttaki H, Stams AJ, Schink B, et al. Physiology, ecology, phylogeny, and genomics of microorganisms capable of syntrophic metabolism. Ann N Y Acad Sci. 2008;1125:58–72.
CAS  Article  Google Scholar 

50.
Sieber J, McInerney M, Plugge C, Schink B, Gunsalus R. Methanogenesis: syntrophic metabolism. In: Timmis KN (ed), Handbook of hydrocarbon and lipid microbiology. Springer: Berlin, Heidelberg; 2010. p. 337–55.

51.
Vandieken V, Mußmann M, Niemann H, Jørgensen BB. Desulfuromonas svalbardensis sp. nov. and Desulfuromusa ferrireducens sp. nov., psychrophilic, Fe(III)-reducing bacteria isolated from Arctic sediments, Svalbard. Int J Syst Evol Microbiol. 2006;56:1133–9.
CAS  Article  Google Scholar 

52.
Jones DL, Edwards AC. Influence of sorption on the biological utilization of two simple carbon substrates. Soil Biol Biochem. 1998;30:1895–902.
CAS  Article  Google Scholar 

53.
Bray MS, Wu J, Reed BC, Kretz CB, Belli KM, Simister RL, et al. Shifting microbial communities sustain multiyear iron reduction and methanogenesis in ferruginous sediment incubations. Geobiology. 2017;15:678–89.
CAS  Article  Google Scholar 

54.
Dolfing J, Tiedje JM. Acetate inhibition of methanogenic, syntrophic benzoate degradation. Appl Environ Microbiol. 1988;54:1871–3.
CAS  PubMed Central  Article  PubMed  Google Scholar 

55.
Warikoo V, McInerney MJ, Robinson JA, Suflita JM. Interspecies acetate transfer influences the extent of anaerobic benzoate degradation by syntrophic consortia. Appl Environ Microbiol. 1996;62:26–32.
CAS  PubMed Central  Article  PubMed  Google Scholar 

56.
Elshahed MS, McInerney MJ. Benzoate Fermentation by the anaerobic bacterium Syntrophus aciditrophicus in the absence of hydrogen-using microorganisms. Appl Environ Microbiol. 2001;67:5520–5.
CAS  PubMed Central  Article  PubMed  Google Scholar 

57.
Watanabe M, Kojima H, Fukui M. Review of Desulfotomaculum species and proposal of the genera Desulfallas gen. nov., Desulfofundulus gen. nov., Desulfofarcimen gen. nov. and Desulfohalotomaculum gen. nov. Int J Syst Evol Microbiol. 2018;68:2891–9.
CAS  Article  Google Scholar 

58.
Harwood CS, Burchhardt G, Herrmann H, Fuchs G. Anaerobic metabolism of aromatic compounds via the benzoyl-CoA pathway. FEMS Microbiol Rev. 1998;22:439–58.
CAS  Article  Google Scholar 

59.
Rabus R, Boll M, Heider J, Meckenstock RU, Buckel W, Einsle O, et al. Anaerobic microbial degradation of hydrocarbons: from enzymatic reactions to the environment. J Mol Microbiol Biotechnol. 2016;26:5–28.
CAS  Google Scholar 

60.
Podosokorskaya OA, Kadnikov VV, Gavrilov SN, Mardanov AV, Merkel AY, Karnachuk OV, et al. Characterization of Melioribacter roseus gen. nov., sp. nov., a novel facultatively anaerobic thermophilic cellulolytic bacterium from the class Ignavibacteria, and a proposal of a novel bacterial phylum Ignavibacteriae. Environ Microbiol. 2013;15:1759–71.
CAS  Article  Google Scholar 

61.
Kadnikov VV, Mardanov AV, Podosokorskaya OA, Gavrilov SN, Kublanov IV, Beletsky AV, et al. Genomic analysis of Melioribacter roseus, facultatively anaerobic organotrophic bacterium representing a novel deep lineage within Bacteriodetes/Chlorobi group. PLoS ONE 8:e53047. https://doi.org/10.1371/journal.pone.0053047.

62.
Zavarzina DG, Sokolova TG, Tourova TP, Chernyh NA, Kostrikina NA, Bonch-Osmolovskaya EA. Thermincola ferriacetica sp. nov., a new anaerobic, thermophilic, facultatively chemolithoautotrophic bacterium capable of dissimilatory Fe(III) reduction. Extremophiles. 2007;11:1–7.
CAS  Article  Google Scholar 

63.
Wrighton KC, Agbo P, Warnecke F, Weber KA, Brodie EL, DeSantis TZ, et al. A novel ecological role of the Firmicutes identified in thermophilic microbial fuel cells. ISME J. 2008;2:1146–56.
CAS  Article  Google Scholar 

64.
Byrne-Bailey KG, Wrighton KC, Melnyk RA, Agbo P, Hazen TC, Coates JD. Complete genome sequence of the electricity-producing “Thermincola potens” strain JR. J Bacteriol. 2010;192:4078–9.
CAS  PubMed Central  Article  PubMed  Google Scholar 

65.
Wrighton KC. Following electron flow: from a gram-positive community to mechanisms of electron transfer. Berkeley, CA, USA: UC Berkeley; 2010.

66.
Poser A, Lohmayer R, Vogt C, Knoeller K, Planer-Friedrich B, Sorokin D, et al. Disproportionation of elemental sulfur by haloalkaliphilic bacteria from soda lakes. Extremophiles. 2013;17:1003–12.
CAS  Article  Google Scholar 

67.
Sorokin DY, Tourova TP, Mußmann M, Muyzer G. Dethiobacter alkaliphilus gen. nov. sp. nov., and Desulfurivibrio alkaliphilus gen. nov. sp. nov.: two novel representatives of reductive sulfur cycle from soda lakes. Extremophiles. 2008;12:431–9.
CAS  Article  Google Scholar 

68.
Zhuang L, Tang Z, Ma J, Yu Z, Wang Y, Tang J. Enhanced anaerobic biodegradation of benzoate under sulfate-reducing conditions with conductive iron-oxides in sediment of Pearl River Estuary. Front Microbiol. 2019;10:374.
PubMed Central  Article  PubMed  Google Scholar 

69.
Kamagata Y, Kitagawa N, Tasaki M, Nakamura K, Mikami E. Degradation of benzoate by an anaerobic consortium and some properties of a hydrogenotrophic methanogen and sulfate-reducing bacterium in the consortium. J Ferment Bioeng. 1992;73:213–8.
CAS  Article  Google Scholar 

70.
Junghare M, Schink B. Desulfoprunum benzoelyticum gen. nov., sp. nov., a gram-negative benzoate-degrading sulfate-reducing bacterium isolated from the wastewater treatment plant. Int J Syst Evol Microbiol. 2015;65:77–84.
CAS  Article  Google Scholar 

71.
Oren A. The order Halanaerobiales, and the families Halanaerobiaceae and Halobacteroidaceae. In: Rosenberg E, DeLong EF, Lory S, Stackebrandt E, Thompson F, editors. The prokaryotes: firmicutes and tenericutes. Berlin, Heidelberg: Springer Berlin Heidelberg; 2014. p. 153–77.

72.
Hatamoto M, Imachi H, Yashiro Y, Ohashi A, Harada H. Detection of active butyrate-degrading microorganisms in methanogenic sludges by RNA-based stable isotope probing. Appl Environ Microbiol. 2008;74:3610–4.
CAS  PubMed Central  Article  PubMed  Google Scholar 

73.
Nobu MK, Narihiro T, Liu M, Kuroda K, Mei R, Liu WT. Thermodynamically diverse syntrophic aromatic compound catabolism. Environ Microbiol. 2017;19:4576–86.
CAS  Article  Google Scholar 

74.
Lentini CJ, Wankel SD, Hansel CM. Enriched iron(III)-reducing bacterial communities are shaped by carbon substrate and iron oxide mineralogy. Front Microbiol. 2012;3:404.
PubMed Central  Article  PubMed  Google Scholar 

75.
Newsome L, Lopez Adams R, Downie HF, Moore KL, Lloyd JR. NanoSIMS imaging of extracellular electron transport processes during microbial iron(III) reduction. FEMS Microbiol Ecol. 2018;94:fiy104.

76.
Wang H, Byrne JM, Liu P, Liu J, Dong X, Lu Y. Redox cycling of Fe(II) and Fe(III) in magnetite accelerates aceticlastic methanogenesis by Methanosarcina mazei. Environ Microbiol Rep. 2020;12:97–109.
CAS  Article  Google Scholar 

77.
Dodsworth JA, Blainey PC, Murugapiran SK, Swingley WD, Ross CA, Tringe SG, et al. Single-cell and metagenomic analyses indicate a fermentative and saccharolytic lifestyle for members of the OP9 lineage. Nat Commun. 2013;4:1854.
PubMed Central  Article  CAS  PubMed  Google Scholar 

78.
Nobu MK, Dodsworth JA, Murugapiran SK, Rinke C, Gies EA, Webster G, et al. Phylogeny and physiology of candidate phylum ‘Atribacteria’(OP9/JS1) inferred from cultivation-independent genomics. ISME J. 2016;10:273–86.
CAS  Article  Google Scholar 

79.
Algora C, Vasileiadis S, Wasmund K, Trevisan M, Krüger M, Puglisi E, et al. Manganese and iron as structuring parameters of microbial communities in Arctic marine sediments from the Baffin Bay. FEMS Microbiol Ecol. 2015;91:fiv056.
Article  CAS  PubMed  PubMed Central  Google Scholar 

80.
Lehours A-C, Rabiet M, Morel-Desrosiers N, Morel J-P, Jouve L, Arbeille B, et al. Ferric iron reduction by fermentative strain BS2 isolated from an iron-rich anoxic environment (Lake Pavin, France). Geomicrobiol J. 2010;27:714–22.
CAS  Article  Google Scholar 

81.
Liu D, Wang H, Dong H, Qiu X, Dong X, Cravotta CA. Mineral transformations associated with goethite reduction by Methanosarcina barkeri. Chem Geol. 2011;288:53–60.
CAS  Article  Google Scholar 

82.
Sivan O, Shusta S, Valentine D. Methanogens rapidly transition from methane production to iron reduction. Geobiology. 2016;14:190–203.
CAS  Article  Google Scholar 

83.
Whitman WB, Coleman DC, Wiebe WJ. Prokaryotes: the unseen majority. Proc Nat Acad Sci USA. 1998;95:6578–83.
CAS  Article  Google Scholar 

84.
Aromokeye AD. Iron oxide driven methanogenesis and methanotrophy in methanic sediments of Helgoland Mud Area, North Sea. Bremen, Germany: Universität Bremen; 2018. More

The methodological framework for simulating the dispersal of bryophytes under changing climate conditions is presented in Fig. 4. A grid of pixel-specific environmental conditions and dispersal kernels, combining information on species dispersal traits, local wind conditions, as well as landscape features affecting dispersal by wind, is generated and used as input in simulations of species dispersal in the landscape under changing climate conditions.
Fig. 4: Overview of workflow implemented in the present study to integrate mechanistic dispersal kernels and correlative climatic suitability models in simulations of future wind-dispersed species distributions under climate change.

Species distribution data (left) are combined with climatic variables to produce climatic suitability models that are calibrated under present and projected under future climatic conditions (Part 1) and used to build mechanistic dispersal models (Part 2). The latter combine species intrinsic features (spore settling velocity Vt and release height Z0) and extrinsic environmental features (mean horizontal wind speed Ū and canopy height h) to generate maps of spatially explicit dispersal kernels. Climatic suitability and dispersal kernel maps, updated at regular intervals, are finally combined to parameterize simulations of dynamic range shifts under changing climatic conditions (Part 3).

Full size image

Data sampling
The European bryophyte flora includes 1817 native or naturalized species41. Because information on bryophyte species distribution is scarce and very heterogeneous, challenging the application of climatic suitability models42, we selected 10 species based upon their representativeness for each of the four main biogeographic elements (i.e., groups of species sharing similar distribution patterns), namely the Arctic-Alpine, Atlantic, Mediterranean, and wide-temperate elements (Supplementary Table 2). For each of these species, we downloaded data from the Global Biodiversity Information Facility (https://www.gbif.org). We excluded data collected before 1960, which represented, on average, 41 ± 12% of the data available, for two reasons. First, old records often lack sufficiently precise location information. Second, we wanted to avoid a potential mismatch between old observations and current climate conditions used for modeling. To complete these data and generate a dataset across the entire range of each species in Europe, we specifically performed a thorough literature review to document their occurrence from more than 600 sources. Only points that were separated by at least 0.1° from each other were subsequently retained for modeling (“ecospat.occ.desaggregation” function in Ecospat 3.143) to avoid sampling bias and reduce the risk of spatial autocorrelation. Altogether, the number of observations available for each species ranged between 55 and 34,035 (database available from Figshare, https://doi.org/10.6084/m9.figshare.8289650).
Average spore diameter was recorded for each species from Zanatta et al.44 and references therein. Species unknown to produce sporophytes were assigned a spore size of 150 µm to take dispersal through larger asexual propagules into account. Spore settling velocities Vt and release height (0.03, 1 and 10 m, which roughly correspond to habitat preferences for ground-dwelling, saxicolous, and epiphytic species, respectively) were determined for each species (Supplementary Table 2) following Zanatta et al.44.
Nineteen bioclimatic variables, averaged over the period from 1970 to 2000, were retrieved from WorldClim 1.4 at a resolution of 30 arc-seconds45. Although snow is an important driver of species distributions in Arctic regions46, the lack of sufficiently detailed information on snow precipitation across Europe prevented us from implementing this variable.
Given the spatial grain of our study, the hypothesis that some species will persist in small microhabitats, where temperatures can be cooler and humidity higher than in the surrounding environment, cannot be rejected. Data at finer scales for both present and future conditions would therefore be desirable47. Recently developed methods to generate fine-grained climatic data taking into account microclimatic effects modulated by microtopographic variation in the terrain, vegetation cover and ground properties using energy balance equations cannot, however, yet be implemented across large spatial scales48.
For future climate conditions, a wide range of GCMs have been described and their variation represents the largest source of uncertainty in future range prediction studies49. No criterion exists to evaluate GCMs, whose performance may vary among regions and variables50. Due to computational constrains associated with our migration simulations (see below), we followed Didersky et al.51. and selected two GCMs that reflected the highest and lowest levels of predicted changes due to climate change for two angiosperm species in Europe50, namely MPI-ESM-LR52 and HadGem2-ES53. For each GCM, we analyzed two climate change scenarios. These scenarios are expressed by the representative concentration pathways (RCPs), using values comparing the level of radiative forcing between the preindustrial era and 2100. The moderate scenario RCP4.5 assumes 650 ppm CO2 and 1.0–2.6 °C increase by 2100, and refers to AR4 guideline scenario B1 of IPCC AR4 guidelines. The pessimistic scenario RCP8.5 assumes 1350 ppm CO2 and 2.6–4.8 °C increase by 2100, and refers to A1F1 scenario of IPCC AR4 guidelines54. Climatic data for each GCM and each RCP were averaged for each of the four time periods considered, i.e., 2010–2020, 2020–2030, 2030–2040 and 2040–2050.
Monthly average and daily maximum wind speeds measured at 10 m as well as predicted wind speeds for the same ten-year time periods between 2010 and 2050, were computed from EURO-CORDEX (https://euro-cordex.net). Canopy height data were obtained from the global scale mapping of canopy height and biomass at a 1-km spatial resolution55. Wind speed and canopy height were sampled for each pixel and each time-slice to generate kernel maps through time (see below).
Deriving climatic suitability maps
The correlation among the 19 bioclimatic variables was computed from 50,000 random points. To avoid multicollinearity, five bioclimatic variables with a Pearson correlation value of R 10 km from a potential source could be colonized by LDD. The maximum LDD distance was set to unlimited based on phylogeographic evidence39. Following Robledo-Arnuncio et al.31, we employed the results of previous Approximate Bayesian Computation methods for LDD inference from genetic structure data in bryophytes39,77 to define the range of LDD probability values, set to 0, 10−4, 10−3, 10−2 and 10−1.
Migclim simulations
We modeled the dispersal of a species under a climate change scenario over a period of 40 years, from 2010 to 2050. Starting with an initial distribution for the year 2010, the climatic suitability of cells was updated every 10 years to reflect the projected changes in climatic conditions under the considered climate change scenario. Since our simulations run over 40 years, we need four different climatic suitability maps. The wind layers were updated at the same 10 years intervals as the climatic data to produce series of spatially and temporally explicit kernel maps. We assume that our species disperse once a year, and hence, our simulations performed a total of 40 dispersal steps between 2010 and 2050. For each 10 years climatic period, pixels were identified as potentially suitable based on the binarized climatic suitability model projections. While climatic suitability thus drove colonization probability, a recent study raised the intriguing idea that spread rates at the migration front increase as climatic suitability decreases as a response to the need to seek for more suitable habitats78. In bryophytes, however, such a mechanism would be unlikely as inadequate resources and investment in environmental stress defence typically result in shifts from sexual to asexual reproduction79.
For each species, we ran a sensitivity analysis by testing the impact of variation of the free parameters described above: two values of horizontal windspeed Ū (monthly average and daily maximum), three values of spore release height Z0 (0.03, 1 and 10 m), and four values of LDD probabilities (see above). For each parameter combination, 30 MigClim replicates were performed.
We computed the ratio between the predicted loss of suitable area (fraction of initially suitable cells that became unsuitable by 2050) and the simulated effective colonization rate (fraction of newly suitable cells by 2050 that were effectively colonized) using two extreme values of the LDD probability range, that is, 0 and 0.1.
To determine the time-lag of the colonization of newly suitable habitats, the analyses were run for 500 years, keeping the environmental parameters at their 2050 values.
Reporting summary
Further information on research design is available in the Nature Research Reporting Summary linked to this article. More

Drosophila rearing
The starting colony of D. suzukii was collected from wild Rubus sp. and Fragaria sp. fruits in various sites in Switzerland in 201523. The flies from the initial collection are described molecularly by Fraimout et al.43. The starting colony of D. melanogaster and D. simulans were obtained from laboratory colonies of INRA (Sophia-Antipolis, France) in 2015 and 2019, respectively. The general rearing of flies was done in plastic tubes (5 cm diameter, 10 cm height) containing approximately 10 g of artificial diet (Formula 4-24 medium, Carolina Biological SupplyCo., Burlington, NC), 40 ml of methyl-4-hydroxylbenzoate solution (1.43 g/L) to inhibit fungal growth, and a few grains of commercial instant dry yeast. The tubes were kept in growth chambers at 22 ± 2 °C, 60% ± 10% RH, and a 16 h photoperiod (hereafter called general rearing conditions). To collect eggs and resulting larvae on different nutritive media (i.e., fresh and decomposing fruits or artificial diet) for the below-described parasitoid rearing and experiments with parasitoids, some adult flies were kept in gauze cages (BugDorm-4F4545) at general rearing conditions. They were fed with sugar water provided on dental cotton rolls and dried instant yeast, additional water was provided on cellulose paper. The nutritive media were exposed to adult flies when needed.
Parasitoid rearing
The starting colonies of G. cf. brasiliensis were obtained during surveys in Asia from 2015–2017 and names to describe their origin are based on the collection sites described by Girod et al.19: Dali, Fumin, Kunming, Shiping, and Kunming—Xining temple (Xining in this study) in the Yunnan Province of China, as well as Hasuike (Nagano) and Tokyo—Naganuma park (actually on the territory of Hachioji but named Tokyo in this study) in Japan. The parasitoids were reared in the quarantine laboratory at CABI-Switzerland (Delémont, Switzerland) separated by origin in gauze cages (BugDorm-4F4545) to prevent them from interbreeding. The general rearing was done on D. suzukii larvae feeding on blueberries as described by Girod et al.19, with the difference that fruits were only exposed for 24 h to D. suzukii for oviposition. The environmental parameters of the quarantine chamber were the above-described general rearing conditions. Up to 50 adult wasps were kept in transparent plastic containers (9 cm diameter, 5 cm height) inside each gauze cage. An Eppendorf tube with a wet cellulose paper was added as a water source and the container was closed with a foam plug on which a drop of honey was placed as food source. Six fresh blueberries, which were placed 24 h before in the D. suzukii rearing cages to collect eggs, were added every 2–3 days to each container with adults to allow for parasitism of young fly larvae. After the exposure to the wasps, infested fruits were removed from the containers and kept in clear plastic tubes (5 cm diameter, 10 cm height) with a filter paper at the bottom to absorb leaking fruit juice. Every 2–3 days, the presence of newly hatched wasps was checked among rearing tubes and adult wasps were transferred to the oviposition containers.
Molecular characterization
The molecular characterization was performed on (1) individuals originating from the field (nine locations from five provinces in China and three locations from three prefectures of the Honshu island in Japan), (2) the derived laboratory strains and (3) individuals used for the experiments (Table S2). Two molecular markers were used, the mitochondrial coding gene Cytochrome Oxidase subunit 1 (COI) and the nuclear region Internal Transcripted Spacer 2 (ITS2). Both were previously used to characterize Ganaspis individuals from Eastern Asia22,23 and elsewhere.
The DNA was extracted in a total of 30 µl using either the prepGEM Insect kit (Zygem) (3 h at 75 °C and 5 min at 95 °C), or the QuickExtract DNA Extraction Solution (n°QE09050, Lucigen) (15 min at 65 °C and 2 min at 98 °C). For both molecular markers (COI and ITS2), each individual PCR was realized in a total of 25 µl, including 12.5 µl of the Multiplex PCR Master Mix (Qiagen), 0.125 µl of each primer (100 µM), and 1 µl DNA. For COI, the primers LCO (5′-GGTCAACAAATCATAAAGATATTGG-3′) and HCO (5′-TAAACTTCAGGGTGACCAAAAAATCA-3′)44 were used for more than 400 individuals. PCR conditions consisted of (1) 15 min at 95 °C, (2) 35 cycles of 30 s at 94 °C, 90 s at 50 °C and 60 s at 72 °C, (3) 10 min at 72 °C. For ITS2, the primers ITS2-F (5′-TGTGAACTGCAGGACACATG-3′) and ITS2-R (5′-AATGCTTAAATTTAGGGGTA-3′)45 were used for a subset of representative individuals. PCR conditions consisted of (1) 15 min at 95 °C; (2) 40 cycles of 30 s at 94 °C, 90 s at 53 °C, and 60 s at 72 °C; and (3) 10 min at 72 °C. In both cases, the PCR was checked using a QIAxcel DNA Fast Analysis Kit on a QIAxcel Advanced System (Qiagen). Positive PCR products were then sequenced with the Sanger method in one direction with the HCO primer for COI and both directions for ITS2. Sequences were trimmed, assembled and aligned using ClustalW for COI and Muscle for ITS2 (Geneious, version 10.2.3). For COI, only haplotypes observed twice within the panel of high-quality sequences (length  > 520 bp and no undetermined nucleotide) were considered. These data were then enriched with 83 additional GenBank accessions, including in particular sequences from Nomano et al.22 and Giorgini et al.23. The whole dataset (our own haplotypes and GenBank accessions) was then analyzed on a common part of 519 bp included between the two marks, ATTGGDTCAA and TTAGCAGGTG (5′ → 3′ on the positive strand). Three criteria were then applied to summarize and clean the data including: (1) the conservation of repres entative, necessary and sufficient sequences from the three main sources22,23 (and this study); (2) the exclusion of sequence with undetermined nucleotide(s); (3) the exclusion of each sequence with a unique amino-acid sequence. A final dataset of 62 sequences (haplotypes from this study and GenBank accessions) remained after this process. Based on this dataset, three complementary approaches were used to investigate the molecular clustering: (1) a Neighbour Joining approach using the Tamura 3 parameters distance (the best evolutionary model according to the software MEGA10.1.746), using 500 replicates for bootstrapping; (2) a Maximum Likelihood approach using the evolutionary model HKY85 + I (the best model according to the software PhyML3.047); and (3) the constitution of a network using the Median Joining method (ε set to zero, PopArt48). The Kimura 2 parameters distance (often used in the frame of barcoding’s studies) was also used to investigate the pairwise distances within and between clusters (see Discussion). For ITS2, the identified haplotypes were directly compared to those available on GenBank and mapped into the COI Neighbor-Joining tree.
Crossing experiments
Ganaspis brasiliensis is arrhenotokous, unmated females produce only male progeny while mated females are able to produce both males (unfertilized eggs) and females (fertilized eggs). Thus, the proportion of female progeny can be used as an indicator of reproductive isolation. With regard to already acquired knowledge on Asian Ganaspis cf. brasiliensis19,22,25,30, we more precisely investigated here the reproductive (in)compatibilities between the two main molecular clusters (G1 and G3-4—see Results and Discussion) and, within the cluster G1, between two geographically distant populations (one Chinese and one Japanese). Thus, crossing experiments with individuals from three locations were done here: Tokyo, Hasuike and Kunming. For the latter, only individuals that were a posteriori affiliated to G1 through the molecular characterization described above were taken into account. For individuals from each location, parasitized Drosophila pupae from the general parasitoid rearing (see above) were identified under a microscope (parasitoid pupae can be seen through the translucent Drosophila pupal case) and kept individually in plastic vials containing moisturized plastic foams. Within 24 h after emergence, 1–2 males were placed with each virgin female during 24 h for mating. Females were then transferred to a plastic vial containing 10–30 first instar D. suzukii larvae feeding in fresh blueberries and drops of honey for the parasitoid’s nutrition. After 3 days, females were collected and kept in 95% ethanol for potential molecular analysis. The vials containing the potentially parasitized D. suzukii larvae in blueberries were kept until adult emergence under the general rearing conditions described above. Upon emergence of the F1 generation, adults were sexed based on antennal length (males have longer antennae than females24) and the percentage of female progeny was calculated for each parental female. To test the fertility of F1 females, they were allowed mating with males from the same origin for 24 h. Then, the above described oviposition procedure was repeated, and upon emergence, the F2 progeny was sexed and percentage of females was calculated. The number of parental females for each crossing varied from 9–24 (Table 1), depending on emergence during the experimental period.
Affinity towards the targeted host and its nutritive media
To study the specificity of G. cf. brasiliensis from the above mentioned seven different origins in Asia, three combinations of hosts and nutritive media were tested under no-choice conditions: (1) D. suzukii larvae feeding on blueberries, (2) D. suzukii larvae feeding on artificial diet, and (3) D. melanogaster larvae feeding on artificial diet. The blue formula of the above-mentioned artificial diet was used to facilitate counting of Drosophila eggs. Additionally, the diet was blended with about 25 g of fresh blueberries, as described by Girod et al.25. The artificial diet and fresh blueberries were exposed to the respective Drosophila species for 1–3 h, until 10–30 eggs were counted under a microscope, and incubated for 24 h at room temperature to allow eggs to hatch. Mated and naïve (i.e., never exposed to hosts for oviposition) 3–4 d old G. cf. brasiliensis females were then released individually into plastic tubes (2.7 cm diameter, 5.2 cm height) containing one of the three media. The tubes were closed with a moist foam lid containing a drop of honey to nourish the parasitoids. Females were removed from the tubes after 48 h and placed in 95% ethanol for genetic identification based on CO1, as described above. The tubes containing potentially parasitized Drosophila larvae were kept at the general rearing conditions and observed for fly and parasitoid emergence on a regular basis for 40 d. For each tube, the number of Drosophila flies and parasitoids were recorded. For each parasitoid origin, 20 replicates per host species-nutritive media combination were tested, for a total of 420 individual females.
Influence of the nutritive media on the parasitism of non-target species
A second no-choice test was done to investigate whether G. cf. brasiliensis’ host specificity is dependent on the nutritive medium of the host. To this end, four host species-nutritive medium combinations were tested: D. melanogaster or D. simulans larvae feeding on either blueberries or artificial diet. Because both Drosophila species do not have a serrated ovipositor and can therefore not oviposit through the skin of fresh fruits, slightly decomposed blueberries were cut in half and exposed to these species until 10–30 eggs were counted on each half. As in the first no-choice test, the artificial diet used in this experiment was the blue formula blended with about 25 g blueberries. The experiment was then conducted as described above for the first no-choice test, with the difference that 10 replicates for each host species-nutritive medium combination were used for parasitoids originating from Tokyo, Xining, and Hasuike only. This brought the total number of females for this experiment to 120.
Preference for the targeted host and its habitats
To investigate differences in preferences for the targeted host and its habitats among the different genetic groups of G. cf. brasiliensis, a three- and a four-choice bioassay were done. The bioassays took place in a cylindrical transparent plastic container (10 cm diameter, 5 cm height) with two holes of 2.5 cm diameter in the lid: one was covered with netting for ventilation and the other closed with a foam plug on which a drop of honey was placed to nourish the parasitoid. Inside each container, one 4–5 days old mated parasitoid female was placed, a plastic vial with wet cellulose paper as a water source, and small dishes (2.5 cm diameter, 1 cm height) containing the choices for oviposition in a random order. To avoid the influence of light and colors on the wasp’s directional choice, the choice arenas were placed inside a white plastic box (100 × 50 cm), leaving only one light source from above. After 24 h in the choice arena at the general rearing conditions, female parasitoids were kept in 95% ethanol to allow for further DNA analysis confirming the genetic group they belonged to. The dishes containing the different hosts and nutritive media were placed separately in rearing tubes (5 cm diameter, 10 cm height) containing a moist filter paper at the bottom and covered with a moist foam lid to avoid drying of the media. Three weeks after the beginning of the choice test, all adult Drosophila were removed from the rearing tubes and were counted. Until the eighth week after the choice test, emerging parasitoids were collected once a week, sexed, and counted.
The three-choice bioassay was designed to determine if also when given the choice, G1 G. cf. brasiliensis are specific to fruits as the host’s nutritive medium, rather than to the host species, while G3-4 parasitoids are not specific to either. Therefore, the three host-species-nutritive medium combinations were (1) D. suzukii or (2) D. melanogaster larvae feeding on fresh blueberry, and (3) D. melanogaster larvae feeding on artificial diet. All media were prepared as described above for the no-choice experiments. In total, 68 female wasps were tested in the three-choice bioassay, 20 originating from Hasuike, 24 from Tokyo, and 24 from Xining.
To determine if the habitat specificity of G1 and generality of G3-4 G. cf. brasiliensis also hold true when comparing fresh to decomposing fruits, a four-choice bioassay was designed. The host species-nutritive media combinations were (1) D. suzukii or (2) D. melanogaster larvae feeding on either (3) fresh or (4) decomposing blueberry. Infestation of fresh blueberries with fly larvae was done as described above. To decompose fruits, blueberries were exposed to room temperature in a plastic container for 7–10 days until growth of molt was visible. They were then exposed to D. suzukii and D. melanogaster for the collection of eggs as described for fresh fruits. In total, 27 and 22 females originating from Tokyo (G1) and Hasuike (G3-4) were tested, respectively, in the four-choice bioassay. For all choice tests, only results from females that produced at least one offspring were analyzed.
Statistical analysis
Apparent parasitism (AP) was calculated as the proportion of parasitoid offspring among the total number of insects that emerged from the nutritive medium (i.e. Drosophila sp. and parasitoids). The proportion of ovipositing females (POF) was calculated as the number of female parasitoids which produced at least one offspring (or which showed an oviposition response, in the case of the behavioral experiments) divided by the number of females tested. All data were analyzed using logistic regression followed by post-hoc comparisons of means with Tukey adjustments. Differences in proportions of females in the crossing experiment as well as AP and POF in the no-choice experiments was analyzed using quasibinomial distributions to account for overdispersion of the residuals (glm function of the ‘stats’ package in R49). For the no-choice experiment with parasitoids from different origins, AP was analyzed with the explanatory variables parasitoid origin, nutritive medium, and their interaction; and the POF developing on D. melanogaster feeding on artificial diet was analyzed with the parasitoid’s genetic group (G1 or G3-4) as explanatory variable. AP in the no-choice experiment with non-target species, the explanatory variables were parasitoid origin, host species, nutritive medium, and all possible interactions.
Mixed effects logistic regressions (glmer function of the ‘lme4’ package in R50) were used to analyze AP in the choice tests. Analyses were done for each parasitoid origin separately because of convergence problems with more than one fixed effect. Therefore, nutritive medium was the sole fixed-effect explanatory variable for all analyses concerning the choice tests. In all cases, individual females were included as a random effect to account for correlation of parasitism between the media by the same female and an additional observation-level random effect was introduced to solve the problem of residual overdispersion. More