Modelling the evolution of complete genome sequences in populations requires accounting for the recombination process, as a single tree can no longer describe the underlying genealogy. The sequentially Markov coalescent (SMC, McVean and Cardin 2005; Marjoram and Wall 2006) approximates the standard coalescent with recombination process and permits estimating population genetic parameters (e.g., population sizes, recombination rates) using population genomic datasets. As such datasets become available for an increasing number of species, more fine-tuned models are needed to encompass the diversity of life cycles of organisms beyond the model species on which most methods have been benchmarked.

The work by Korfmann et al. (Korfmann et al. 2024) represents a significant step forward as it accounts for multiple mergers in SMC models. Multiple merger models account for simultaneous coalescence events so that more than two lineages find a common ancestor in a given generation. This feature is not allowed in standard coalescent models and may result from selection or skewed offspring distributions, conditions likely met by a broad range of species, particularly microbial.

Yet, this work goes beyond extending the SMC, as it introduces several methodological innovations. The "classical" SMC-based inference approaches rely on hidden Markov models to compute the likelihood of the data while efficiently integrating over the possible ancestral recombination graphs (ARG). Following other recent works (e.g. Gattepaille et al. 2016), Korfmann et al. propose to separate the ARG inference from model parameter estimation under maximum likelihood (ML). They introduce a procedure where the ARG is first reconstructed from the data and then taken as input in the model fitting step. While this approach does not permit accounting for the uncertainty in the ARG reconstruction (which is typically large), it potentially allows for the extraction of more information from the ARG, such as the occurrence of multiple merging events. Going away from maximum likelihood inference, the authors trained a graph neural network (GNN) on simulated ARGs, introducing a new, flexible way to estimate population genomic parameters.

The authors used simulations under a beta-coalescent model with diverse demographic scenarios and showed that the ML and GNN approaches introduced can reliably recover the simulated parameter values. They further show that when the true ARG is given as input, the GNN outperforms the ML approach, demonstrating its promising power as ARG reconstruction methods improve. In particular, they showed that trained GNNs can disentangle the effects of selective sweeps and skewed offspring distributions while inferring past population size changes.

This work paves the way for new, exciting applications, though many questions must be answered. How frequent are multiple mergers? As the authors showed that these events "erase" the record of past demographic events, how many genomes are needed to conduct reliable inference, and can the methods computationally cope with the resulting (potentially large) amounts of required data? This is particularly intriguing as micro-organisms, prone to strong selection and skewed offspring distributions, also tend to carry smaller genomes.

References

Gattepaille L, Günther T, Jakobsson M. 2016. Inferring Past Effective Population Size from Distributions of Coalescent Times. Genetics 204:1191-1206.
https://doi.org/10.1534/genetics.115.185058
 
Korfmann K, Sellinger T, Freund F, Fumagalli M, Tellier A. 2024. Simultaneous Inference of Past Demography and Selection from the Ancestral Recombination Graph under the Beta Coalescent. bioRxiv, 2022.09.28.508873. ver. 5 peer-reviewed and recommended by Peer Community in Evolutionary Biology. https://doi.org/10.1101/2022.09.28.508873
 
Marjoram P, Wall JD. 2006. Fast "coalescent" simulation. BMC Genet. 7:16.
https://doi.org/10.1186/1471-2156-7-16
 
McVean GAT, Cardin NJ. 2005. Approximating the coalescent with recombination. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 360:1387-1393.
https://doi.org/10.1098/rstb.2005.1673

The genome of eukaryotic species is a complex structure that experience many different interactions within itself and with the surrounding environment. The genetic architecture of a phenotype (that is, the set of genetic elements affecting a trait of the organism) plays a fundamental role in understanding the adaptation process of a species to, for example, different climate environments, or to its interaction with other species. Thus, it is fundamental to study the different aspects of the genetic architecture of the species and its relationship with its surronding environment. Aspects such as modularity (the number of genetic units and the degree to which each unit is affecting a trait of the organism), pleiotropy (the number of different effects that a genetic unit can have on an organism) or linkage (the degree of association between the different genetic units) are essential to understand the genetic architecture and to interpret the effects of selection on the genome. Indeed, the knowledge of the different aspects of the genetic architecture could clarify whether genes are affected by multiple aspects of the environment or, on the contrary, are affected by only specific aspects [1,2].

The work performed by Lotterhos et al. [3] sought to understand the genetic architecture of the adaptation to different environments in lodgepole pine (Pinus contorta), considering as candidate SNPs those previously detected as a result of its extreme association patterns to different environmental variables or to extreme population differentiation. This consideration is very important because the study is only relevant if the studied markers are under the effect of selection. Otherwise, the genetic architecture of the adaptation to different environments would be masked by other (neutral) kind of associations that would be difficult to interpret [4,5]. In order to understand the relationship between genetic architecture and adaptation, it is relevant to detect the association networks of the candidate SNPs with climate variables (a way to measure modularity) and if these SNPs (and loci) are affected by single or multiple environments (a way to measure pleiotropy).

The authors used co-association networks, an innovative approach in this field, to analyse the interaction between the environmental information and the genetic polymorphism of each individual. This methodology is more appropriate than other multivariate methods - such as analysis based on principal components - because it is possible to cluster SNPs based on associations with similar environmental variables. In this sense, the co-association networks allowed to both study the genetic and physical linkage between different co-associations modules but also to compare two different models of evolution: a Modular environmental response architecture (specific genes are affected by specific aspects of the environment) or a Universal pleiotropic environmental response architecture (all genes are affected by all aspects of the environment). The representation of different correlations between allelic frequency and environmental factors (named galaxy biplots) are especially informative to understand the effect of the different clusters on specific aspects of the environment (for example, the co-association network ‘Aridity’ shows strong associations with hot/wet versus cold/dry environments).

The analysis performed by Lotterhos et al. [3], although it has some unavoidable limitations (e.g., only extreme candidate SNPs are selected, limiting the results to the stronger effects; the genetic and physical map is incomplete in this species), includes relevant results and also implements new methodologies in the field. To highlight some of them: the preponderance of a Modular environmental response architecture (evolution in separated modules), the detection of physical linkage among SNPs that are co-associated with different aspects of the environment (which was unexpected a priori), the implementation of co-association networks and galaxy biplots to see the effect of modularity and pleiotropy on different aspects of environment. Finally, this work contains remarkable introductory Figures and Tables explaining unambiguously the main concepts [6] included in this study. This work can be treated as a starting point for many other future studies in the field.

References

[1] Hancock AM, Brachi B, Faure N, Horton MW, Jarymowycz LB, Sperone FG, Toomajian C, Roux F & Bergelson J. 2011. Adaptation to climate across the Arabidopsis thaliana genome. Science 334: 83–86. doi: 10.1126/science.1209244
[2] Wagner GP & Zhang J. The pleiotropic structure of the genotype­phenotype map: the evolvability of complex organisms. Nature Review Genetics 12: 204–213. doi: 10.1038/nrg2949
[3] Lotterhos KE, Yeaman S, Degner J, Aitken S, Hodgins K. 2018. Modularity of genes involved in local adaptation to climate despite physical linkage. bioRxiv 202481, ver. 4 peer-reviewed by Peer Community In Evolutionary Biology. doi: 10.1101/202481
[4] Lotterhos KE & Whitlock MC. 2014. Evaluation of demographic history and neutral parameterization on the performance of FST outlier tests. Molecular Ecology 23: 2178–2192. doi: 10.1111/mec.12725
[5] Lotterhos KE & Whitlock MC. 2015. The relative power of genome scans to detect local adaptation depends on sampling design and statistical method. Molecular Ecology 24: 1031–1046. doi: 10.1111/mec.13100
[6] Paaby AB & Rockman MV. 2013. The many faces of pleiotropy. Trends in Genetics 29: 66-73. doi: 10.1016/j.tig.2012.10.010

The observed levels of genomic diversity in contemporary populations are the result of changes imposed by several evolutionary processes. Among them, natural selection is known to dramatically shape the genetic diversity of loci associated with phenotypes which affect the fitness of carriers. As such, many efforts have been dedicated towards developing methods to detect signatures of natural selection from genomes of contemporary samples [1].
Recent technological advances made the generation of large-scale genomic data from temporal samples, either from experimental populations or historical or ancient samples, accessible to a wide scientific community [2]. Notably, temporal population genomic data allow for a direct observation and study of how, for instance, allele frequencies change through time in response to evolutionary stimuli. Such information can be exploited to detect loci under natural selection, either via mathematical modelling or by investigating empirical distributions [3].
However, most of current methods to detect selection from temporal genomic data have largely ignored selfing populations, despite the latter comprising a significant proportion of species with social and economic importance. Selfing changes genomic patterns by reducing the effective recombination rate, which makes distinguishing between neutral evolution and natural selection even more challenging than for the case of outcrossing populations [4]. Nevertheless, an outlier-approach based on temporal genomic data for the selfing Arabidopsis thaliana population revealed loci under selection [5].
This study suggested the promise of detecting selection for selfing populations and encouraged further investigations to test the power of selection scans under different mating systems.
To address this question, Navascués et al. [6] extended a previously proposed approach for temporal genome scan [7] to incorporate partial self-fertilization. In the original implementation [7], it is assumed that, under neutrality, all loci provide levels of genetic differentiation drawn from the same distribution. If some of the loci are under selection, such distribution should show heterogeneity. Navascués et al. [6] proposed a test for the homogeneity between loci-specific and genome-wide differentiation by deriving a null distribution of FST via simulations using SLiM [8]. After filtering for low-frequency variants and correct for multiple tests, authors derived a statistical test for selection and assess its power under a wide range of scenarios of selfing rate, selection coefficient, duration and type of selection [6].
The newly proposed test achieved good performance to distinguish between neutral and selected loci in most tested scenarios.
As expected, the test's performance significantly drops for scenarios of high selfing rates and selection from standing variation. Additionally, the probability to correctly detect selection decreases with increasing distance from the causal variant. Intriguingly, the test showed high power when the selected ancestral allele had an initial low frequency, and when the selected derived allele had a high initial frequency. When applied to a data set of around 1,000 SNPs from the highly selfing Medicago truncatula population, an annual plant of the legume family [9], the test did not provide any candidate loci under selection [6].
In summary, the detection of loci under selection in selfing populations is and largely remains a challenging task even when explictly account for the different mating system. However, recombination events that occurred before the selective pressure allow ancestral beneficial alleles to exhibit a detectable pattern of non-neutrality. As such, in partially selfing populations, the strength of the footprint of selection depends on several factors, mostly on the selfing rate, the time of onset and type of selection.
One major assumption of this study is that the model implies unstructured population and continuity between samples obtained from the same geographical location over time. As such assumptions are typically violated in real populations, further research into the effect of more complex demographic scenarios is desired to fully understand the power to detect selection in selfing populations. Furthermore, more power could be gained by including additional genomic information at each time point. In this context, recent approaches that make full use of genomic data based on deep learning [10] may contribute significantly towards this goal. Similarly, the effect of data filtering on the power to detect selection should be further explored, especially in the context of DNA resequencing experiments. These analyses will help elucidate the power offered by selection scans from temporal genomic data in selfing populations.

References

[1] Stern AJ, Nielsen R (2019) Detecting Natural Selection. In: Handbook of Statistical Genomics , pp. 397–40. John Wiley and Sons, Ltd. https://doi.org/10.1002/9781119487845.ch14
[2] Leonardi M, Librado P, Der Sarkissian C, Schubert M, Alfarhan AH, Alquraishi SA, Al-Rasheid KAS, Gamba C, Willerslev E, Orlando L (2017) Evolutionary Patterns and Processes: Lessons from Ancient DNA. Systematic Biology, 66, e1–e29. https://doi.org/10.1093/sysbio/syw059
[3] Dehasque M, Ávila‐Arcos MC, Díez‐del‐Molino D, Fumagalli M, Guschanski K, Lorenzen ED, Malaspinas A-S, Marques‐Bonet T, Martin MD, Murray GGR, Papadopulos AST, Therkildsen NO, Wegmann D, Dalén L, Foote AD (2020) Inference of natural selection from ancient DNA. Evolution Letters, 4, 94–108. https://doi.org/10.1002/evl3.165
[4] Vitalis R, Couvet D (2001) Two-locus identity probabilities and identity disequilibrium in a partially selfing subdivided population. Genetics Research, 77, 67–81. https://doi.org/10.1017/S0016672300004833
[5] Frachon L, Libourel C, Villoutreix R, Carrère S, Glorieux C, Huard-Chauveau C, Navascués M, Gay L, Vitalis R, Baron E, Amsellem L, Bouchez O, Vidal M, Le Corre V, Roby D, Bergelson J, Roux F (2017) Intermediate degrees of synergistic pleiotropy drive adaptive evolution in ecological time. Nature Ecology and Evolution, 1, 1551–1561. https://doi.org/10.1038/s41559-017-0297-1
[6] Navascués M, Becheler A, Gay L, Ronfort J, Loridon K, Vitalis R (2020) Power and limits of selection genome scans on temporal data from a selfing population. bioRxiv, 2020.05.06.080895, ver. 4 peer-reviewed and recommended by PCI Evol Biol. https://doi.org/10.1101/2020.05.06.080895
[7] Goldringer I, Bataillon T (2004) On the Distribution of Temporal Variations in Allele Frequency: Consequences for the Estimation of Effective Population Size and the Detection of Loci Undergoing Selection. Genetics, 168, 563–568. https://doi.org/10.1534/genetics.103.025908
[8] Messer PW (2013) SLiM: Simulating Evolution with Selection and Linkage. Genetics, 194, 1037–1039. https://doi.org/10.1534/genetics.113.152181
[9] Siol M, Prosperi JM, Bonnin I, Ronfort J (2008) How multilocus genotypic pattern helps to understand the history of selfing populations: a case study in Medicago truncatula. Heredity, 100, 517–525. https://doi.org/10.1038/hdy.2008.5
[10] Sanchez T, Cury J, Charpiat G, Jay F Deep learning for population size history inference: Design, comparison and combination with approximate Bayesian computation. Molecular Ecology Resources, n/a. https://doi.org/10.1111/1755-0998.13224

The ability of a population to adapt to a new niche is an important phenomenon in evolutionary biology. The colonisation of a new volcanic island by plant species; the colonisation of a host treated by antibiotics by a-resistant strain; the Ebola virus transmitting from bats to humans and spreading epidemically in Western Africa, are all examples of a population invading a new niche, adapting and eventually establishing in this new environment.

Adaptation to a new niche can be studied using source-sink models. In the original environment —the “source”—, the population enjoys a positive growth-rate and is self-sustaining, while in the new environment —the “sink”— the population has a negative growth rate and is able to sustain only by the continuous influx of migrants from the source. Understanding the dynamics of adaptation to the sink environment is challenging from a theoretical standpoint, because it requires modelling the demography of the sink as well as the transient dynamics of adaptation. Moreover, local selection in the sink and immigration from the source create distributions of genotypes that complicate the use of many common mathematical approaches.

In their paper, Lavigne et al. [1], develop a new deterministic model of adaptation to a harsh sink environment in an asexual species. The fitness of an individual is maximal when a number of phenotypes are tuned to an optimal value, and declines monotonously as phenotypes are further away from this optimum. This model —called Fisher’s Geometric Model— generates a GxE interaction for fitness because the phenotypic optimum in the sink environment is distinct from that in the source environment [2]. The authors circumvent mathematical difficulties by developing an original approach based on tracking the deterministic dynamics of the cumulant generating function of the fitness distribution in the sink. They derive a number of important results on the dynamics of adaptation to the sink:

• From the point where immigration from the source to the sink starts, four phases of adaptation are observed. After a short transient phase (phase 1), a migration-selection balance is reached in the sink (phase 2). After a while, thanks to the immigration of rare adapted migrants and mutation in the sink, a small fraction of the sink population exhibits a close-to-optimal phenotype. This small adapted fraction grows in frequency and mean fitness rapidly increases in the sink (phase 3). Finally, the population settles around the sink optimum (phase 4) and, hurray, the sink is now a source!

• Interestingly, in this model the evolutionary dynamics do not depend on the immigration rate. In other words, adaptation will proceed at the same rate regardless of how many immigrants invade the sink. This is because the impact of immigration on adaptation depends on the rate of immigration relative to the sink density. This ratio is actually independent of immigration in a model where the sink is initially empty, migration from the sink back to the source is negligible and without density-dependence in the sink.

• In this model, mutation is a double-edged sword. Adapted phenotypes emerge from new mutations, and under this effect alone a higher mutation rate would translate into a shorter time to establishment in the sink. However, mutations may also have deleterious effects by displacing the phenotype away from the optimum. This mutation load will be greater when individuals need to simultaneously tune a large number of phenotypes. As a consequence of these two effects of mutations, time to establishment is minimal for an intermediate mutation rate. This result emerges from Fisher’s Geometric Model, but may hold more generally for biologically plausible fitness landscapes where mutations generates both beneficial (allowing adaptation to the sink) and deleterious genotypes.

• Lastly, in Fisher’s Geometric Model, the time to establishment increases superlinearly with harshness of the sink when the sink is too harsh, and establishment may occur only after a very long time. In these harsh sinks, the adapted genotypes are very few and increase very slowly in frequency, making the second phase of adaptation much longer. Thus, and as a direct consequence of Fisher’s Geometric Model, adding a “stepping stone” intermediate environment would allow faster adaptation to the extreme environment.

In conclusion, this theoretical work presents a method based on Fisher’s Geometric Model and the use of cumulant generating functions to resolve some aspects of adaptation to a sink environment. It generates a number of theoretical predictions for the adaptive colonisation of a sink by an asexual species with some standing genetic variation. It will be a fascinating task to examine whether these predictions hold in experimental evolution systems: will we observe the four phases of the dynamics of mean fitness in the sink environment? Will the rate of adaptation indeed be independent of the immigration rate? Is there an optimal rate of mutation for adaptation to the sink? Such critical tests of the theory will greatly improve our understanding of adaptation to novel environments.

References

[1] Lavigne, F., Martin, G., Anciaux, Y., Papaïx, J., and Roques, L. (2019). When sinks become sources: adaptive colonization in asexuals. bioRxiv, 433235, ver. 5 peer-reviewed and recommended by PCI Evolutionary Biology. doi: 10.1101/433235
[2] Martin, G., and Lenormand, T. (2006). A general multivariate extension of Fisher's geometrical model and the distribution of mutation fitness effects across species. Evolution, 60, 893-907. doi: 10.1111/j.0014-3820.2006.tb01169.x

Domesticated organisms are excellent models for understanding ecology and evolution and they are important for our food production and safety. While less studied than plants and animals, micro-organisms have also been domesticated, in particular for food fermentation [1]. The most studied domesticated micro-organism is the yeast used to make wine, beer and bread, Saccharomyces cerevisiae [2, 3, 4].

Filamentous fungi used for cheese-making have recently gained interest, for example Penicillium roqueforti used to make blue cheeses and P. camemberti to make soft cheeses [5, 6, 7, 8]. As for plants and animals, domestication has led to beneficial traits for food production in fermenting fungi, but also to bottlenecks and degeneration [6, 7, 9]; P. camemberti for example does not produce enough spores any more for optimal culture and inoculation and P. roqueforti has lost sexual fertility [9]. The loss of genetic diversity and of species diversity in our food production system is concerning for multiple reasons : i) it jeopardizes future improvement in the face of global changes ; ii) it causes the loss of evolved diversity during centuries under human selection, and therefore of beneficial characteristics and specificities that we may never be able to recover ; iii) it leads to degeneration in the few cultivated strains; iv) it impoverishes the diversity of our food products and local adaptation of production practices. 

The study of domesticated fungi used for food fermentation has focused so far on the evolution of lineages and on their metabolic specificities. Microbiological assemblages and species diversity have been much less studied, while they likely also have a strong impact on the quality and safety of final products. This study by Elisa Michel and colleagues [10] addresses this question, using an interdisciplinary participatory research approach including bakers, psycho-sociologists and microbiologists to analyse bread-making practices and their impact on microbial communities in sourdough. Elisa Michel and colleagues [10] identified two distinct groups of bread-making practices based on interviews and surveys, with farmer-like practices (low bread production, use of ancient wheat populations, manual kneading, working at ambient temperature, long fermentation periods and no use of commercial baker’s yeast) versus more intensive, artisanal-like practices. Metabarcoding and microbial culture-based analyses showed that the well-known baker’s yeast, Saccharomyces cerevisiae, was, surprisingly, not the most common species in French organic sourdoughs. Kazachstania was the most represented yeast genus over all sourdoughs, both in terms of read abundance and of species diversity. Kazachstania species were also often dominant in individual sourdoughs, but Saccharomyces uvarum or Torulaspora delbrueckii could also be the dominant yeast species.

Metabarcoding analyses further revealed that the composition of the fungal communities differed between the farmer-like and more intensive practices, representing the first evidence of the influence of artisanal practices on microbial communities. The fungal communities were impacted by a combination of bread-making variables including the type of wheat varieties, the length of fermentation, the quantity of bread made per week and the use of commercial yeast. Maintaining on farm less intensive bread-making practices, may allow the preservation of typical species and phenotypic diversity in microbial communities in sourdough. Farmer-like practices did not lead to higher diversity within sourdoughs but, overall, the diversity of bread-making practices allow maintaining a larger diversity in sourdoughs. For example, different Kazachstania species were most abundant in sourdoughs from artisanal-like and farmer-like practices. Interviews with the bakers suggested the role of dispersal of Kazachstania species in shaping sourdough microbial communities, dispersal occurring by seed exchanges, sourdough mixing or gifts, bread-making training in common or working in one another’s bakery. Nikolai Vavilov [11] had already highlighted for crops the importance of isolated cultures and selection in different farms for generating and preserving crop diversity, but also the importance of seed exchange for fostering adaptation. 

Furthermore, one of the yeast frequently found in artisanal sourdoughs, Kazachstania humilis, displayed phenotypic differences between sourdough and non-sourdough strains, suggesting domestication. The sourdough strains exhibited significantly higher CO2 production rate and a lower fermentation latency-phase time. 

The study by Elisa Michel and colleagues [10] is thus novel and inspiring in showing the importance of interdisciplinary studies, combining metabarcoding, microbiology and interviews for assessing the composition and diversity of microbial communities in human-made food, and in revealing the impact of artisanal-like bread-making practices in preserving microbial community diversity.

Interdisciplinary studies are still rare but have already shown the importance of combining ethno-ecology, biology and evolution to decipher the role of human practices on genetic diversity in crops, animals and food microorganisms and to help preserving genetic resources [12]. For example, in the case of the bread wheat Triticum aestivum, such interdisciplinary studies have shown that genetic diversity has been shaped by farmers’ seed diffusion and farming practices [13]. We need more of such interdisciplinary studies on the impact of farmer versus industrial agricultural and food-making practices as we urgently need to preserve the diversity of micro-organisms used in food production that we are losing at a rapid pace [6, 7, 14]. 

References

[1] Dupont J, Dequin S, Giraud T, Le Tacon F, Marsit S, Ropars J, Richard F, Selosse M-A (2017) Fungi as a Source of Food. Microbiology Spectrum, 5, 5.3.09. https://doi.org/10.1128/microbiolspec.FUNK-0030-2016

[2] Legras J-L, Galeote V, Bigey F, Camarasa C, Marsit S, Nidelet T, Sanchez I, Couloux A, Guy J, Franco-Duarte R, Marcet-Houben M, Gabaldon T, Schuller D, Sampaio JP, Dequin S (2018) Adaptation of S. cerevisiae to Fermented Food Environments Reveals Remarkable Genome Plasticity and the Footprints of Domestication. Molecular Biology and Evolution, 35, 1712–1727. https://doi.org/10.1093/molbev/msy066

[3] Bai F-Y, Han D-Y, Duan S-F, Wang Q-M (2022) The Ecology and Evolution of the Baker’s Yeast Saccharomyces cerevisiae. Genes, 13, 230. https://doi.org/10.3390/genes13020230

[4] Fay JC, Benavides JA (2005) Evidence for Domesticated and Wild Populations of Saccharomyces cerevisiae. PLOS Genetics, 1, e5. https://doi.org/10.1371/journal.pgen.0010005

[5] Ropars J, Rodríguez de la Vega RC, López-Villavicencio M, Gouzy J, Sallet E, Dumas É, Lacoste S, Debuchy R, Dupont J, Branca A, Giraud T (2015) Adaptive Horizontal Gene Transfers between Multiple Cheese-Associated Fungi. Current Biology, 25, 2562–2569. https://doi.org/10.1016/j.cub.2015.08.025

[6] Dumas E, Feurtey A, Rodríguez de la Vega RC, Le Prieur S, Snirc A, Coton M, Thierry A, Coton E, Le Piver M, Roueyre D, Ropars J, Branca A, Giraud T (2020) Independent domestication events in the blue-cheese fungus Penicillium roqueforti. Molecular Ecology, 29, 2639–2660. https://doi.org/10.1111/mec.15359

[7] Ropars J, Didiot E, Rodríguez de la Vega RC, Bennetot B, Coton M, Poirier E, Coton E, Snirc A, Le Prieur S, Giraud T (2020) Domestication of the Emblematic White Cheese-Making Fungus Penicillium camemberti and Its Diversification into Two Varieties. Current Biology, 30, 4441-4453.e4. https://doi.org/10.1016/j.cub.2020.08.082

[8] Caron T, Piver ML, Péron A-C, Lieben P, Lavigne R, Brunel S, Roueyre D, Place M, Bonnarme P, Giraud T, Branca A, Landaud S, Chassard C (2021) Strong effect of Penicillium roqueforti populations on volatile and metabolic compounds responsible for aromas, flavor and texture in blue cheeses. International Journal of Food Microbiology, 354, 109174. https://doi.org/10.1016/j.ijfoodmicro.2021.109174

[9] Ropars J, Lo Y-C, Dumas E, Snirc A, Begerow D, Rollnik T, Lacoste S, Dupont J, Giraud T, López-Villavicencio M (2016) Fertility depression among cheese-making Penicillium roqueforti strains suggests degeneration during domestication. Evolution, 70, 2099–2109. https://doi.org/10.1111/evo.13015

[10] Michel E, Masson E, Bubbendorf S, Lapicque L, Nidelet T, Segond D, Guézenec S, Marlin T, Devillers H, Rué O, Onno B, Legrand J, Sicard D, Bakers TP (2022) Artisanal and farmer bread making practices differently shape fungal species community composition in French sourdoughs. bioRxiv, 679472, ver. 6 peer-reviewed and recommended by Peer Community in Evolutionary Biology. https://doi.org/10.1101/679472

[11] Vavilov NI, Vavylov MI, Dorofeev VF (1992) Origin and Geography of Cultivated Plants. Cambridge University Press.

[12] Saslis-Lagoudakis CH, Clarke AC (2013) Ethnobiology: the missing link in ecology and evolution. Trends in Ecology & Evolution, 28, 67–68. https://doi.org/10.1016/j.tree.2012.10.017

[13] Thomas M, Demeulenaere E, Dawson JC, Khan AR, Galic N, Jouanne-Pin S, Remoue C, Bonneuil C, Goldringer I (2012) On-farm dynamic management of genetic diversity: the impact of seed diffusions and seed saving practices on a population-variety of bread wheat. Evolutionary Applications, 5, 779–795. https://doi.org/10.1111/j.1752-4571.2012.00257.x

[14] Demeulenaere É, Lagrola M (2021) Des indicateurs pour accompagner “ les éleveurs de microbes” : Une communauté épistémique face au problème des laits “ paucimicrobiens ” dans la production fromagère au lait cru (1995-2015). Revue d’anthropologie des connaissances, 15. http://journals.openedition.org/rac/24953