Rising together: Exploring Sourdough Fermentation Diversity through Co-design in the HealthFerm Citizen Science Initiative

Rising together: Exploring Sourdough Fermentation Diversity through Co-design in the HealthFerm Citizen Science Initiative | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Rising together: Exploring Sourdough Fermentation Diversity through Co-design in the HealthFerm Citizen Science Initiative Annina Meyer, Thomas Gettemans, Jan Patrick Tan, Fabio Tuccillo, and 14 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6627352/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 27 Jan, 2026 Read the published version in npj Science of Food → Version 1 posted You are reading this latest preprint version Abstract Fermented foods are culturally significant and increasingly recognized for their potential health benefits, yet scientific data on household fermentation practices remain limited. We launched a co-designed citizen science (CS) initiative within the HealthFerm project to collect sourdough fermentation data across Europe. Over 1000 participants from 33 countries registered, with 671 samples submitted, enabling large-scale analysis of fermentation practices, motivations, and sourdough characteristics. Participants also completed standardized at-home experiments and sensory evaluations, generating a dataset linking baking habits with physicochemical and sensory profiles. Distinct patterns emerged: professional bakers used older, more frequently refreshed starters and fermented at higher temperatures. Ingredient choices and motivations varied by country, shaped by perceived health benefits. Beyond data collection, this initiative established a microbial biobank and harmonized metadata resource, while offering practical insights into co-design, logistics, and public engagement. The resulting framework provides a transferable model for participatory research in microbiology and food systems science. Social science/Science technology and society Biological sciences/Microbiology/Applied microbiology Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Introduction Fermented foods, produced through desired microbial growth and enzymatic activities (Marco et al., 2021 ), are the primary natural source of microorganisms in the human diet, and hold significant cultural, nutritional, and ecological values (Hernández-Velázquez et al., 2024 ). Fermented foods are popularly perceived as beneficial for health, possibly due to the nature of fermentation as a traditional, low-input food processing technique, and source of dietary microorganisms. However, despite their roles in shaping fermented food products, supporting regional biodiversity, influencing gut health (Valentino et al., 2024 ), and even contributing to planetary well-being (Cavicchioli et al., 2019 ; Gupta et al., 2016 ), the functional significance of microorganisms in food production and health remains largely overlooked by the general public. Studying food microbiomes presents significant hurdles, particularly in obtaining samples that reflect diverse geographical, cultural, and procedural origins, a limitation compounded by the constrained reach of traditional research approaches. Citizen science (CS) offers a transformative solution to overcome these challenges while also strengthening the connection between scientists and the public. Through public involvement in microbiome research — ranging from simple contributory efforts to fully co-designed initiatives — scientists can gather more representative datasets that reflect the ecological, cultural, culinary, or health-related diversity of microbiomes. Beyond expanding the scope of research, CS fosters community participation in scientific discovery, mutual learning and creates awareness of the critical roles microorganisms play in daily life (Garcia et al., 2023 ). It also acts as a platform for innovation, public engagement, and evidence-based policy advocacy, ensuring the microbial world’s significance is recognised and valued across disciplines and societies (Fraisl et al., 2022 ). Integrating CS into system-level microbiome studies unlocks insights essential for addressing global challenges such as antimicrobial resistance, climate change, and food security (Lewandowski et al., 2017 ; Pateman et al., 2020 ). In fermented food research, it also highlights overlooked microbial ecosystems and deepens public and academic understanding of microbial diversity, bridges scientific research with traditional knowledge, and underscores its role in sustainable food systems, cultural heritage, and human and planetary health. For example, Kefir4All and the Global Sourdough Project highlighted the roles of microbial biodiversity in shaping the production and quality of home-made fermented foods (Landis et al., 2021 ; Walsh et al., 2024 ). Through extensive public participation, diverse outreach, and transparent dissemination of findings, these projects showcase the power of CS to generate robust datasets, challenge prevailing assumptions, and strengthen connections between scientists and the public. Sourdough, a traditional fermented mixture of water and cereal flour(s) with roots traceable to ancient times, has played a key role in the evolution of diverse food cultures by enhancing the sensory attributes, texture, nutritional value, and microbiological safety of cereal-based foods (Cappelle et al., 2013 ). Fermentation allows microbial growth and metabolism in the dough, producing sourdough that can serve as a leavening agent and can be propagated through backslopping—adding a portion of fermented dough to fresh flour and water (Arora et al., 2021 ; Calvert et al., 2021 ; De Vuyst et al., 2017 , 2023 ). Despite its simple ingredients, traditional sourdoughs exhibit remarkable microbial variability, influenced by differences in household and bakery practices, flour type, dough yield, backslopping ratio, and fermentation/storage conditions (Comasio et al., 2020 ; De Vuyst et al., 2017 ; Landis et al., 2021 ). As most studies to date focus on laboratory-prepared or small-scale bakery sourdoughs, comprehensive large-scale investigations of spontaneously fermented (Type 1) sourdoughs in real-world settings are still needed (Bondt et al., 2024 ). Here, we describe the design and execution of a large-scale sourdough fermentation study, utilizing a CS co-design approach to facilitate greater collaboration with participants. The study was conducted within the framework of the HealthFerm project ( https://healthferm.eu/ ) , a European initiative exploring innovative pulse- and cereal-based food fermentations, their potential health effects, and consumer perceptions. Through its citizen science component, HealthFerm aims to advance understanding of the sourdough ecosystem and develop novel microbial resources to support the targeted redesign of fermented foods, promoting healthier and more sustainable diets. Here, we highlight important study design considerations as well as the technical challenges and potential biases inherent to this approach and present sourdough citizen science insights. Key outcomes include broad participant engagement, the creation of geographically diverse datasets, and valuable insights for future CS initiatives. Altogether, the project explored sourdough maintenance practices and consumer preferences and perceptions across Europe, establishing a framework for microbiologists designing collaborative, volunteer-driven microbiome research. Results Collaborative, iterative co-design guides methodological citizen science implementation and refinement Collaborative methodological planning was carried out with HealthFerm partner institutions and through direct engagement with citizens. Five universities served as local citizen science centers (hubs) to promote the study, recruit participants, and collect sourdough samples (Fig. S1 a). Three hubs organized co-design workshops located in Herent (Belgium), Helsinki (Finland), and Zurich (Switzerland), involving 70 participants (27, 17, and 26, respectively). These workshops played a central role in shaping and refining methodologies throughout the project (Fig. 1 , Fig. S1 b). Participants showed strong engagement and willingness to conduct at-home experiments and complete follow-up questionnaires. Their feedback informed improvements to the sampling kits, such as simplifying terminology, adding a project overview, and clarifying return instructions. Participant questions and interests were gathered to guide dissemination across project stages (Supplementary list 1). Participants from different countries reported varying practices for sourdough refreshment, particularly in recipe precision. In Helsinki, only 66% used scales and exact recipes, compared to 85% in Zurich, reflecting a preference for volumetric units in Finland. Therefore, to standardize at-home measurements, a volume-to-mass conversion table was included in the instruction booklet. Feedback also guided decisions on sample submission and communication preferences. Local postal services were selected as the main delivery method, with optional drop-off for those near hub labs; sampling kits were distributed exclusively by mail. In response to requests for centralized updates, a multilingual blog was added to the HealthFerm website (Fig. S1 b). Engagement patterns and sample contributions in a trans-European citizen science study Extensive outreach across Europe led to strong engagement primarily from European as well as some non-European participants (Fig. 2 a). By September 2024, the project had registered 1091 sourdoughs and collected 671 samples from 591 participants across 33 countries, including 27 in Europe (Fig. 2 b). Most sign-ups (68.6%) and samples (65.7%) originated from hub countries (Switzerland, Finland, Belgium, Italy, Romania) and Germany (Fig. 2 c), resulting in a slightly skewed language distribution (Fig. 2 d). 61 individuals submitted more than one sourdough (n = 2–5). Recruitment success varied by country and channel (Fig. 2 e, Fig. S2a): sourdough blogs were particularly effective in Germany, press in Switzerland, LinkedIn in Denmark, and Facebook in Finland (Table S2). In contrast, word-of-mouth in Finland and Facebook in Switzerland and Germany were less impactful. Merging sign-up forms and survey responses showed that 87.2% of sourdoughs came from household bakers (Fig. 2 f), with 67.7% self-initiated and 8.0% passed on by others (Fig. 2 g). Regarding experience, 26.6% had baked for under two years, while over a third had five or more years of experience or were professionals (Fig. 2 h). Most sourdoughs were less than a year old at sampling (Fig. 2 i), with a notable peak of newly started sourdoughs in 2020, reflecting the surge of popularity in sourdough baking during the COVID-19 pandemic; the oldest registered sourdough dated back to 1865. Fermentation routines and storage conditions differ between households and bakeries Differences in environmental and fermentation practices were observed between professional (n = 109) and home bakers (n = 1019) (Fig. 3 ). Sourdoughs from both groups were proportionally distributed across European countries, minimizing regional bias (Fig. S2b). Professional bakers more often maintained older starters (55% > 5 years vs. 28% in households), refreshed them daily (41.3% vs. 43.7% weekly in households), and fermented at higher temperatures (23°C vs. 21°C) with shorter fermentation times. Both groups typically stored sourdoughs at or below 10°C when not in use. Plastic containers were more common in bakeries (61.5%), while home bakers preferred glass jars (only 9.4% use plastic containers). The presence of plants, pets, or children in the baking environment did not differ significantly, likely due to unclear survey wording. Ingredients were similar across groups (Fig. 3 c): wheat was most common (52.3% bakeries; 53.4% households), followed by rye, with widespread use of organic flours (66.1% bakeries; 59.6% households). Most reported dough yields (DY; \(\:\frac{Flour\:\left(g\right)\:+\:water\:\left(g\right)}{Flour\:\left(g\right)}*100\) ) between 150 and 250, indicating firm to semi-liquid sourdoughs. Baking motivations and perceived health benefits shape ingredient choices Personal motivations and beliefs about sourdough varied between participants. Participants most frequently cited hedonistic motivations such as taste and a general interest in baking, and financial/tradition motivations less frequently (Fig. 4 a). A majority (84.2%) perceived sourdough as healthier than yeast-leavened bread, with no significant difference across experience levels (Fig. S3a, p = 0.115). Gastrointestinal benefits were the most reported perceived health benefit, again consistent across levels of baking experience (Fig. 4 b, Fig. S3b, p = 0.115). Health-motivated bakers significantly more frequently used rye and less frequently wheat as the grain base (Fig. 4 c, Table S3). Tradition-driven bakers more often used rye and spelt, while rye was also overrepresented among those citing health-related reasons such as using "healthier ingredients" or experiencing "fewer intolerances" compared to baking with conventional yeast (Fig. 4 c, Table S3). Bakers using wholemeal or flour mixes more often reported enjoyment, health motivation, and gastrointestinal benefits, compared to those using endosperm or unspecified flour types, who were less likely to cite enjoyment, taste, or digestibility (Fig. 4 c, Table S3). Health-motivated bakers also used organic flour significantly more often and associated it with improved nutritional quality or bioavailability (Fig. 4 c, Table S3). Citizen science experiments reveal patterns and challenges in sourdough physicochemical and sensory characteristics Among participants who submitted sourdough samples, 63.6% (427/671) completed five optional home experiments, while 15.6% did not perform any. In 13.8% of cases, no sourdough bread was baked, and 8.6% of those who baked did not complete the experiments. pH measurements performed at home revealed a positive correlation between sourdough and bread pH (Fig. 5 a), indicating that sourdough acidification generally yielded more acidic bread, regardless of flour type. Bread density, however, was not affected by pH (Fig. 5 a, Fig. S4). Laboratory-measured pH values were on average 0.73 units lower than home measurements (Fig. 5 b, Fig. S5), likely due to further acidification during shipping and/or the use of accurate laboratory pH meters versus pH strips at home. Participants characterised the aroma of their sourdoughs using the Check-All-That-Apply (CATA) methodology and selected from 56 predefined descriptors, organised into 13 categories (Fig. 5 c). Fermented, grain/cereal, fermented dairy, fruity, and sour aromas were most prevalent; ocean, animal, and body odour were least common. High variability in aroma perception was evident in Jaccard dissimilarity-based PCoA (Fig. 5 d), where the first two components explained only 13.7% of the total variance. Features contributing most to the variance included yeasty, bread, and yoghurt all positively correlated with PC1, whereas porridge, hay, lactic acid, alcoholic, vinegar and wholegrain aromas also ranked in the top ten features but negatively correlated with PC1 (Fig. 5 d). PERMANOVA identified categorical variables such as plant presence, organic flour use, milling grade, and acidic beverage fermentation as significantly associated with aroma variation (BH-FDR p < 0.05), although effect sizes were low (R² < 0.005; Table S4). Among numerical variables, wheat proportion and the number of maintained sourdoughs (positive correlation) and backslopping frequency (negative correlation), were significant but weak predictors (Mantel Pearson R < 0.1; Table S5). Notably, not clear separation was observed when grouping samples by flour type in the full aroma feature space (Fig. 5 d, Table S5). Nonetheless, specific aroma descriptors in isolation correlated with experimental data and flour types. Sourdoughs described as “sour” had significantly higher TTAs (Fig. 5 f, Table S6) and were more frequently composed of wholemeal rye flour (Fig. 5 g). Wheat sourdoughs demonstrated distinct aroma characteristics (Fig. 5 g), were less sour and more often associated with chemical odours, which negatively correlated with TTA (Fig. 5 f, Table S6). Substrate-dependent variation in sourdough composition, maintenance, and physicochemical properties The cereal used for sourdough influences its physicochemical properties and associated recipes. Wheat and rye, the most commonly used cereals for sourdough feeding (Fig. 3 c), were selected for in-depth analysis. Reflecting rye’s higher water absorption, participants using rye maintained sourdoughs at higher dough yield than those using wheat (r = 0.30, p = 3.10×10⁻¹⁸; Fig. 6 a). Rye sourdoughs also had slightly lower participant-measured pH (r = 0.12, p = 8.84×10⁻⁴) and higher TTA values (r = 0.37, p = 3.0×10⁻²⁶), despite similar lab pH (p = 0.15) (Fig. 6 b,c). Participants baking with rye reported denser breads than those using wheat (r = 0.37, p = 3.48×10⁻¹¹; Fig. 6 d). These differences were mirrored in sourdough clustering based on Pearson correlations between proportion of flour type used per sourdough and physicochemical or feeding parameters (Fig. 6 e). Sourdoughs with high rye content (wholemeal or not) formed a distinct cluster, associated with higher TTA and dough yield, and lower pH and backslopping frequency. Sourdoughs with non-rye wholemeal flours or high spelt content showed intermediate profiles between rye- and wheat-based sourdoughs (Fig. 6 e; Table S7). Fermentation fingerprints: how region, substrate, and role define sourdough fermentation practices Sourdough ingredient choices varied regionally across Europe, shaped by cultural traditions, climate, and lifestyle. Chi-square enrichment analysis revealed significant country-specific differences in flour types, milling grades, water sources, added ingredients, flour switching, organic use, and concurrent yeast usage (Fig. 7 a, Table S8). Clustering countries by directional Cramér’s V effect sizes highlighted a distinct group of German-speaking countries (Germany, Austria, Switzerland), characterized by a preference for organic wholemeal flours — especially rye and spelt — whereas Romanian, Finnish, and Italian participants more often preferred non-organic, wheat-based sourdoughs. Northwestern European bakers (Belgium, Denmark, Sweden, UK) commonly used mixed flours and grain blends. Water type and commercial yeast practices also varied by country (Fig. 7 a). Regional differences also emerged in baking motivations and perceived health benefits (Fig. 7 b, Table S9), sourdough storage practices, and household environments (Fig. S6, Table S10). Swiss participants emphasised sustainability, Romanians cited health benefits (digestibility, gut health, glucose control), and Italians rarely mentioned health as a motivation. Enjoyment and taste dominate across countries but are less reported in Belgium. Hierarchical clustering shows shared motivational patterns and sourdough baking experience levels between Swiss and Finnish bakers, while Germany and Italy cluster by a higher proportion of experienced bakers and fewer novices (Fig. 7 b). In addition to sourdough, other food and beverage fermentations are commonly practiced in France, Austria, and Germany (especially dairy), but rare in Finland (Fig. S6). Despite patterns related to country of origin and flour preferences, PCA of fermentation features did not yield clustering by country or flour type (Fig. 7 c,d). Instead, PCA consistently separated household and bakery sourdoughs, pointing in opposite directions in both geographic and flour-related feature spaces (Fig. 7 c,d). This household-bakery split was mainly driven by higher backslopping frequency, storage temperature, dough size, age, and fermented dairy aromas in bakery sourdoughs (Fig. 7 c,d). As expected, feeding temperature was associated with shorter fermentation time and dough yield (Fig. 7 c). Together, these results underline both country- and flour-related trends in how sourdoughs are fermented, while emphasizing individual variation and a strong divide between household and bakery sourdoughs. Discussion The diversity of processing practices and parameters used in the production of traditional food fermentations plays a significant, yet often understudied, role in their microbiological and chemical properties, environmental sustainability, cultural heritage, and human health effects. Studying the diversity of home fermentation practices at the continental scale presents a non-trivial challenge for conventional scientific research that typically lacks the means to gather such comprehensive data. Citizen Science (CS) offers a solution to this challenge by enabling large-scale sample collection while simultaneously fostering public engagement and enthusiasm for science, strengthening the connection between scientists and the public, and raising awareness about the importance of scientific research related to food fermentation. The present study demonstrates the potential of a co-design-driven CS approach to investigate the diversity of technological and cultural parameters of sourdough fermentation across Europe. Through the participation of over 1000 citizen scientists and the collection of more than 670 sourdough samples, a large and geographically diverse dataset was established that would have been difficult to obtain solely through conventional research methods. The co-design-based approach facilitated broad public participation and mutual learning between citizens and researchers, resulting in unique insights into sourdough maintenance motivations and practices, ingredient choices, and perceived health benefits. This approach enabled a comprehensive assessment of sourdough fermentation practices across Europe and permitted validation or reconsideration of prevailing assumptions surrounding traditional sourdough fermentation practices. Furthermore, the project established a sourdough biobank that will serve as a valuable resource for future fermented food research. A key distinction between this study and previous initiatives — including CS studies like the Global Sourdough Project and experimental studies like Reese et al . (Landis et al., 2021 ; Reese et al., 2020 ) — is the participatory nature of the HealthFerm citizen science model. Here, participants actively contributed to experiment selection and protocol refinement through feedback loops, making the process more adaptive and participant-centered. While earlier initiatives compiled metadata, the dataset presented in this study includes novel and uniquely detailed information - not only on sourdough fermentation itself, but also on behavioral and cultural dimensions, incorporating comprehensive sensory evaluations and perceived health benefits. Moreover, this co-designed citizen science framework is scalable and transferable to other traditional food systems. It can be tailored to address specific challenges in sample collection or metadata design, underscoring the value of iterative participant engagement in shaping robust research protocols. Pre-launch workshops enhanced engagement by helping to refine sampling kits, instructions, and questionnaires, improving clarity and accessibility. Direct feedback and interactions between citizens and researchers revealed strong participant motivation to contribute to the study, carry out home experiments, and stay informed about project progress and outcomes. These exchanges suggested that hands-on involvement fostered a sense of being valued, sparked curiosity, and strengthened commitment. The collaborative, iterative process appeared to build trust and a sense of ownership, likely sustaining participant interest — especially in the study outcomes. Future citizen science projects may benefit from emphasizing co-design and transparent researcher–participant communication to maintain engagement and foster long-term scientific investment. A significant finding of this study is understanding the public perceptions of the health benefits of sourdough bread. Most participants considered sourdough to be healthier than yeast-leavened bread, commonly citing improved digestibility and gut health as motivating factors for maintaining and baking with sourdough. Public perceptions of health benefits align well with recent scientific findings demonstrating increased mineral bioavailability in sourdough-based products in vitro , and evidence for improved digestibility or lowering the glycaemic index in vivo (Arora et al., 2021 ; D’Amico et al., 2023 ; Ribet et al., 2023 ; Rizzello et al., 2019 ). Nevertheless, these perceptions reflect the intersection of traditional food fermentation, consumer beliefs, and scientific inquiry, emphasizing the relevance of sourdough in public discussions on health and nutrition. The study also faced several challenges. Recruitment biases — particularly self-selection bias, where volunteers were likely more motivated or interested in fermentation than the general public — may have skewed participation. Such imbalances are inherent to citizen science, despite broad outreach, and were compounded by logistical challenges, variability in citizen-conducted experiments, and unpredictable sample arrivals, complicating scheduling of microbiological analyses (Berghen et al., 2024 ; Constant & Hughes, 2023 ; Waugh et al., 2023 ). Geographical coverage also presented challenges. Advertisement strategies across a multilingual continent needed persistence and a multifaceted approach, working best when conducted in local languages. Project recruitment and materials were made available in seven languages, reflecting the national languages of the hub labs involved in this project as well as English. Nevertheless, regional language barriers still posed significant obstacles throughout the recruitment, data collection, and communication phases. This most likely explains the disproportionate geographic representation in the project. Countries sharing a language with the research hubs exhibited a high degree of participation (e.g., Switzerland, Germany, and Austria all share German as one of their official languages) whereas countries without language representation in the project yielded low or no participation (e.g., Spanish, Greek, Polish, etc. ). Future citizen science initiatives may benefit from a strong, multilingual social media presence — potentially supported by digital marketing professionals and enhanced with interactive tools such as chatbots — to increase visibility, foster engagement, and encourage continued participant involvement. Logistical barriers also impacted the study. Sampling kit and sourdough sample shipping was most efficient within the European Union (EU), but international shipments crossing non-EU borders encountered issues related to customs regulations and inconsistencies between postal services (availability of prepaid return labels and other services). Additionally, the absence of temperature-controlled shipping resulted in continued fermentation during transit, and prolonged transit times often led to gas buildup, causing some sourdoughs to leak from their containers. To mitigate this issue in future studies, it may be advisable to instruct participants explicitly to ship samples early in the week, thereby reducing the risk of extended exposure to non-cooled conditions during prolonged transit times (e.g., days when postal delivery services are not available). This continued fermentation possibly explains discrepancies between pH measurements taken at home and those recorded upon arrival at the laboratory. Alternatively, this discrepancy could have arisen from the use of colorimetric pH strips with low dynamic range (blue and green for pH 3 to 6) and less precision than the calibrated electronic pH meters used in the hub labs. Follow-up studies are necessary to evaluate potential microbiological and chemical changes that could have arisen from prolonged fermentation without temperature control. Addressing this uncertainty will help refine protocols for sample collection and shipping, ensuring higher data quality in future CS projects where samples are shipped through the postal system. Despite participant-obtained data often being of similar quality than scientist-obtained data (Lewandowski et al., 2017 ), the large variability in participant-reported data posed significant challenges for curation and interpretation, particularly regarding home experiment results. This variability aligns with findings from citizen science projects, where differences in volunteers’ skills, motivations, and task familiarity can lead to inconsistencies in data collection (Weigelhofer & Pölz, 2016 ; Wiggins & Wilbanks, 2019 ). Sensory evaluations were conducted using the CATA methodology, a reproducible and consumer-appropriate approach to characterize complex aromas. Here, CATA enabled the collection of large volumes and diverse data. However, the observed variability was high - stemming from both sourdough heterogeneity and individual perceptual differences - which underscores the need for cautious interpretation of citizen-contributed sensory data. Extending workshop formats beyond co-design to include practical sessions prior to performing home experiments could enhance data accuracy and reliability in future CS studies (Fleming et al., 2015 ; Piochi et al., 2021 ). Beyond methodological challenges and lessons learned for future CS initiatives, the study revealed valuable insights into sourdough maintenance practices across Europe. Distinct differences emerged between household and bakery sourdoughs, with bakeries typically maintaining older sourdoughs, backslopping more frequently, and fermenting at higher temperatures to meet the fast-paced demands of professional baking. In addition, the choice of cereal also emerged as a key factor shaping sourdough properties. Rye sourdoughs typically exhibited higher dough yields and total titratable acidity compared to wheat-based sourdoughs. These differences can likely be attributed to the higher arabinoxylan content and enzymatic activity of rye flour, which influences water absorption and fermentation kinetics (Courtin & Delcour, 2002 ; Deleu et al., 2020 ). Interestingly, the widespread use of organic flour among many participants reflects a strong consumer preference, suggesting that ingredient choices may influence microbial composition in ways that warrant further investigation in follow-up studies. One of the most intriguing findings of the study was the emergence of strong regional signatures, where countries exhibited distinct ingredient preferences, fermentation habits, and environmental conditions for maintaining sourdough, as well as differences in baking motivations and perceived health benefits. For instance, German-speaking countries (Germany, Austria, Switzerland) showed a clear inclination toward organic wholemeal flours, particularly rye and spelt, reflecting a robust tradition of non-wheat-based bread making. In contrast, participants from southern and northern Europe primarily used wheat-based sourdoughs and non-organic flours, despite their rich sourdough bread cultures. Interestingly, Finnish participants preferred wheat sourdoughs, despite rye bread’s national prominence. This may suggest that rye breads are more commonly purchased than homemade, or it might reflect a latent recruitment bias, where traditional bakers might have been underrepresented or less likely to volunteer for the study. Instead, participants were more likely to be motivated by enjoyment, taste and perceived health benefits rather than tradition, as supported by the fact that fewer than 10% reported baking for traditional reasons, while over 80% cited taste and 70% the enjoyment of baking (Fig. 4 a). This study showcases the potential of co-designed citizen science to enable large-scale, multicultural research and generate comprehensive, population-level datasets that can be integrated with microbiome analyses. It also highlights the need for improved strategies to mitigate demographic biases, ensure data reliability, and address logistical challenges, particularly in multilingual and multinational contexts. The widespread belief in sourdough-related health benefits underscores the importance of continued scientific validation, especially regarding gut microbiota and metabolic health (Wiggins & Wilbanks, 2019 ). Community-generated insights from this study lay a foundation for future investigations into links between sourdough consumption and health outcomes. Moreover, the framework established here offers a scalable model for participatory research, applicable not only to sourdough but also to other traditionally fermented foods. By leveraging citizen participation, future efforts can bridge scientific inquiry with traditional knowledge, advancing our understanding of fermented food microbiomes and their implications for human and planetary health. Materials and methods Implementation and coordination of a multi-hub citizen science study A co-design-driven CS project embedded in the European HealthFerm project lies at the heart of the current study. From the design phase onward, several key considerations were identified and addressed. These include compliance with key privacy and ethical regulations (such as the General Data Protection Regulation (GDPR), the Nagoya Protocol on Access and Benefit Sharing, and ethical approval), the need for communication in the native languages of the citizen scientists throughout the recruitment, experimental, and dissemination phases, logistical challenges such as sample shipping ( e.g. , shipping time, administration, and costs), and unpredictable sample collection timelines influencing the following laboratory processing workload. To manage these complexities, the CS effort was subdivided across five sampling hubs across Europe, each covering a specific geographical area (Fig. 1 ). The different hubs were each responsible for the GDPR and Nagoya protocol compliance within their collection areas. This subdivision also simplified translation efforts, communication with the participants in seven languages, and the shipping arrangements with local couriers. However, it also introduced challenges in ensuring result comparability, requiring the standardisation of protocols and documentation across all hubs. The sampling hubs were located at the Free University of Bozen-Bolzano (UBZ, Italy), Vrije Universiteit Brussel (VUB, Belgium), Institute of Biology Bucharest (IBB, Romania), University of Helsinki (UH, Finland), and ETH Zurich (ETHZ, Switzerland). Co-design workshops and questionnaire design We chose to follow a co-design process to involve representative participants in the design of research objectives and procedures (Fig. 1 , Fig S1 b). These were further refined through multidisciplinary collaboration within the HealthFerm consortium, drawing on expertise in sourdough research, citizen science, microbiome science, food technology, food biochemistry, consumer science, and sensory analysis. To support the co-design process, we prepared draft materials essential for the campaign, including a sourdough metadata questionnaire, an instruction booklet for sampling and at-home experiments, and a sampling kit. Feedback on these prototypes was collected through three co-design workshops held in Belgium, Finland, and Switzerland, involving small groups of citizen scientists representative of the target audience - namely home and small-scale commercial sourdough bakers. All workshops followed a similar structure, beginning with an introductory presentation on the HealthFerm project and the CS initiative, followed by an interactive, layman-friendly presentation on sourdough and the current state of sourdough research, designed to both assess participants’ knowledge levels and to provide them with a foundational understanding of the science behind sourdough. This was followed by an interactive session where participants were guided through the draft questionnaire and instructions, while also trying some proposals for at-home experiments, including a pH measurement and a sensory analysis of a provided sourdough starter. Feedback from the workshop participants was collected throughout the session. The responses obtained were compared across the different workshops and key insights were incorporated in the further development of the sampling kits and sourdough metadata questionnaire used during the subsequent sampling campaign. Participant registration and at-home experimentation The citizen science (CS) sourdough campaign was launched via an online sign-up form (Fig. 1 , Fig. S1 b), accessed through a multilingual registration page on the HealthFerm website (available in English and the six official languages of the participating hub countries). Participants were automatically assigned to a local hub laboratory based on location. Each hub hosted its own language-specific, GDPR-compliant registration survey - including informed consent, pseudonymization, and a sample rights transfer agreement - using Microsoft Forms or Google Forms, depending on institutional preferences. The survey included 90 questions, collecting only essential personal information (e.g., name, email, address) for kit delivery and metadata on the participant’s sourdough. Respondents could also indicate willingness to contribute sourdough or other fermented food samples. Recruitment was promoted via social media, press coverage, print and digital ads, university networks, public lectures, and conferences. Outreach also targeted sourdough communities, fermentation networks, professional bakers, and influencers. A dedicated HealthFerm CS Blog ( https://healthferm.eu/news-and-events/blog ) and FAQ ( https://healthferm.eu/healthferm- community/faq ) were maintained on the project website, with regular updates via LinkedIn and newsletters. To ensure diverse geographic and substrate representation, participants were sub-selected for sample submission. Selected participants received a sampling kit by mail containing instructions, materials for sourdough sampling and home experiments (Table S1 ), and a unique five-character code for pseudonymization. A QR code linked to a follow-up survey (using the same secure tools) to upload experiment results and updated metadata. Participants followed a standardized protocol (archived on Zenodo; see Data availability) to analyze their own sourdoughs and breads. Sourdough pH was measured by mixing 10 g of dough with 10 g of water and using pH strips (MilliPoreSigma, Darmstadt, Germany). Bread pH was assessed by homogenizing 10 g of crumb (crust removed) in 100 g of water. Bread density was determined by measuring a ~ 3 cm crustless slice’s weight and dimensions. Aroma descriptions of sourdough and bread were recorded using sensory wheels developed by the University of Copenhagen for the Wild Sourdough and Smag for Livet projects ( Lav Din Egen Surdej – Mål Og Vurdér | Smagforlivet.Dk , n.d.; Nichols, n.d.) Sourdough sample collection and processing Sourdough samples were sent in ambient conditions by mail for lab analyses. Upon reception, samples were processed within 24 hours at the local collection hubs using common protocols and chemicals. The pH was determined by diluting 10 g of sourdough in 100 mL of deionised water. The same suspension was used to assess the total titratable acidity (TTA) by titrating with 0.1 M NaOH until a pH of 8.5 was reached, with the volume of NaOH added representing the TTA value. For future analyses and biobanking, 5 g of sample was diluted in either 5 g of sterile deionised water or 5 g of sterile 50% glycerol, aliquoted into cryovials, and stored at -80°C. Survey processing, statistics, and data visualization Each hub maintained three multilingual datasets: registration surveys, result surveys, and laboratory measurements. After sample collection, pseudo-anonymised datasets were centrally compiled, translated into English (semi-automated, using Google Sheets), and merged into unified registration, result, and laboratory tables. Manual curation was used to standardise vocabulary and harmonise free-text entries. Datasets were linked via unique 5-character sample codes and person IDs (archived on Zenodo; see Data availability); missing result variables were filled from registration data when available. Some metadata - such as sensory attributes, pH, and bread density - were only collected during the sample phase and are thus available for a subset of samples. Flour types were classified from participant-reported grain and milling grade. In this study, Triticum aestivum is referred to as "wheat," Triticum spelta as "spelt," and Secale cereale as "rye". For chi-square analyses, semi-wholemeal and wholemeal flours were grouped (as semi-wholemeal was only available for wheat and clustered closely with wholemeal wheat in both lab and feeding-parameter data). For selected enrichment analyses, composite flour types (e.g., wholemeal wheat) were decomposed into grain base and milling grade to distinguish associations with participant region, motivation, or health perception. In contrast, numerical and feature-space analyses retained these as separate categories to preserve resolution. Descriptive and correlative analyses excluded missing data on a per-variable basis unless stated otherwise. Data integration, curation, and analysis were conducted using Python 3.12.7 and R 4.4.1 (R Core Team, 2021). Descriptive analyses . Descriptive analysis based on a bakery-household distinction was performed using the dplyr, tidyr, and ggplot2 packages in R. Data were filtered to exclude the samples with an unknown sample source, along with one single sample of industrial origin. Categorisation was introduced for several of the used metadata (dough age, fermentation temperature, fermentation time, storage temperature, and dough yield) by which values present at ≤ 1% prevalence in each category were binned as ‘Other’. Pairwise analysis . Both t-tests and the Mann-Whitney U tests were used depending on the data’s normality and homoscedasticity, while chi-square tests were applied for binary yes/no responses and categorical variables. Effect sizes were calculated using the formula for Cohen’s d, rank serial correlation, and Cramér’s V for t-tests, Mann-Whitney U test, and chi-square tests, respectively. Maps. To visualise the geographic distribution of sourdough samples, spatial data processing and mapping were conducted using Python (v3.10.14), incorporating geopandas (v1.0.1) for geospatial data handling, shapely (v2.0.6) for geometric operations, matplotlib (v3.8.4) for data visualization, and contextility (v1.6.2) for basemap integration. Collinearity analysis. Collinearity among variables was assessed prior to multivariate and enrichment analysis. Categorical associations were tested using Chi-square and Cramér’s V; numerical variables using Pearson, Spearman, and cosine similarity; and mixed types with Kruskal–Wallis tests. False discovery rate was controlled using the Benjamini–Hochberg method. Hierarchical clustering of association matrices guided the selection of representative variables to avoid redundancy and inclusion of collinear variables. Enrichment analyses by category . Due to higher sampling density and geographic representation in Europe, geolocation-based enrichment analyses excluded non-European samples (longitude − 50 to 70, latitude 0 to 70). To assess categorical enrichment across geographic clusters (e.g., countries), Chi-square tests were performed for each category–cluster pair. For significant associations, Cramér’s V quantified effect size, and enrichment scores indicated over- or underrepresentation. P-values were adjusted using Benjamini–Hochberg false discovery rate (BH-FDR). For numerical features, Mann–Whitney U tests were applied to each feature–cluster pair, comparing within-cluster values to all others. Enrichment scores were derived from mean differences, with a small constant (ϵ = 1×10⁻⁴) added to avoid division by zero. Scores were log₂-normalised for scale-independent comparisons. BH-FDR-adjusted p-values were used to classify features as uniquely or multiply enriched. Visualisation followed the same clustering strategy, with only significant log₂-normalised enrichments displayed. For both categorical and numerical enrichments, positive enrichments (PE) and negative enrichments (NE) and uniquely (U) or multiply (M) associated features with clusters were flagged in output tables (Table S2; S3; S8-10) and visualised accordingly. $$\:Category\:Enrichment\:Score\:=\:\frac{Count\:of\:category\:in\:cluster/Total\:count\:in\:cluster}{Total\:count\:of\:category/Total\:samples}$$ $$\:Numerical\:Enrichment\:Score\:=\:\frac{Variable\:mean\:in\:cluster\:+\:\epsilon\:}{Variable\:mean\:overall\:+\:\epsilon\:}$$ Analysis of sourdough aromas. Sensory diversity was assessed using Jaccard distances (full aroma spectrum). PERMANOVA was applied to the distance matrix to test categorical metadata effects, including only variables with ≥ 3 samples and applying BH-FDR correction. Relationships between sensory dissimilarities and numerical metadata were evaluated using Mantel tests (Pearson and Spearman) on scaled variables, with dynamic exclusion of missing data. Principal Coordinate Analysis (PCoA) was performed for visualisation, retaining the first two axes (PC1, PC2). The top 10 aroma features explaining variance were identified via Pearson correlations with principal coordinates. Significant metadata were integrated into the PCoA plot, with scaled feature loadings shown as arrows. Analysis of numerical metadata feature space. Principal Component Analysis (PCA) was performed using scikit-learn. Only non-collinear variables were considered; samples with missing values for the selected variables were excluded. PCA biplots were generated in matplotlib with confidence ellipses added at 95%. Declarations Acknowledgements The authors acknowledge financial support from the project HealthFerm, which is funded by the European Union under the Horizon Europe grant agreement No. 101060247 and by the Swiss State Secretariat for Education, Research and Innovation (SERI) under contract No. 22.00210. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union nor European Research Executive Agency (REA). Neither the European Union nor REA can be held responsible for them. The authors thank Alessandro Rearte, Rosy Mondardini, Wannes De Man, Michael Bom Frøst, Vimac Nolla Ardèvol, Jean-Paul Garin, Elena Martucci, and the rest of the HealthFerm consortium for their advice and support in setting up, promoting and advertising the project. We also thank Luisa Ferreira, Doriela Grabocka, Celine Verdonck, Silvia-Simona Grosu-Tudor, and Emanuela-Catalina Ionetic for their help in data curation and/or technical support. We offer our gratitude to all the participants who attended the co-design workshops, completed the questionnaires, submitted their sourdoughs, performed home experiments, and patiently waited for their results. None of this would have been possible without your contributions. Author information Contributions Conceptualization: N.A.B., S.W., C.M.C., and R.C. conceived and designed the study with input from other authors. Methodology: A.M., T.G., J.P.T., F.C., Y.D.B., and N.A.B. designed the survey and experimental methodology with input from other authors. Investigation: A.M., T.G., J.P.T., C.V., I.-R.A., M.N., and A.Z.A.T. carried out data collection. Curation: A.M. and J.P.T. curated and managed the data. Formal analysis: A.M., J.P.T., and T.G. Visualization: A.M., T.G., and J.P.T. Supervision: N.A.B., L.D., M.G., M.Z., R.C., L.N., S.W., and C.M.C. Writing — original draft: A.M., T.G., J.P.T., and N.A.B. All authors contributed to the review and editing of the manuscript and approved the final version. Ethics declarations This study obtained ethical approval for the activities of each respective hub lab. Approval was granted by the ethics committees of the respective universities under the reference numbers HealthFerm_Cod2023_14 (UniBZ) , ECHW_429 (VUB), 183/2022 (IBB), 52/2023 (UH), and EK 2023-N-118 (ETHZ). Competing interests The authors declare no competing interests. Data availability The dataset supporting this study — including questionnaire files (Survey 1 and Survey 2), the home experiment instruction booklet, and the fully integrated metadata — is currently under restricted access to preserve the integrity of peer review. It will be made publicly available under Creative Commons Attribution 4.0 International (CC BY 4.0) license upon article publication. In the meantime, it is available upon request via Zenodo: https://doi.org/10.5281/zenodo.15366921. Code availability Not applicable. References Arora, K., Ameur, H., Polo, A., Di Cagno, R., Rizzello, C. G., & Gobbetti, M. (2021). Thirty years of knowledge on sourdough fermentation: A systematic review. Trends in Food Science & Technology , 108 , 71–83. https://doi.org/10.1016/j.tifs.2020.12.008 Berghen, B. V., Vanermen, I., & Vranken, L. (2024). Citizen scientists: Unveiling motivations and characteristics influencing initial and sustained participation in an agricultural project. PLOS ONE , 19 (5), e0303103. https://doi.org/10.1371/journal.pone.0303103 Bondt, Y. D., Verdonck, C., Brandt, M. J., Vuyst, L. D., Gänzle, M. G., Gobbetti, M., Zannini, E., & Courtin, C. M. (2024). Wheat Sourdough Breadmaking: A Scoping Review. Annual Review of Food Science and Technology , 15 (Volume 15, 2024), 265–282. https://doi.org/10.1146/annurev-food-110923-034834 Calvert, M. D., Madden, A. A., Nichols, L. M., Haddad, N. M., Lahne, J., Dunn, R. R., & McKenney, E. A. (2021). A review of sourdough starters: Ecology, practices, and sensory quality with applications for baking and recommendations for future research. PeerJ , 9 , e11389. https://doi.org/10.7717/peerj.11389 Cappelle, S., Guylaine, L., & Gobbetti, M. (2013). History and Social Aspects of Sourdough. In Handbook on Sourdough Biotechnology (pp. 1–10). https://doi.org/10.1007/978-1-4614-5425-0_1 Cavicchioli, R., Ripple, W. J., Timmis, K. N., Azam, F., Bakken, L. R., Baylis, M., Behrenfeld, M. J., Boetius, A., Boyd, P. W., Classen, A. T., Crowther, T. W., Danovaro, R., Foreman, C. M., Huisman, J., Hutchins, D. A., Jansson, J. K., Karl, D. M., Koskella, B., Mark Welch, D. B., … Webster, N. S. (2019). Scientists’ warning to humanity: Microorganisms and climate change. Nature Reviews Microbiology , 17 (9), 569–586. https://doi.org/10.1038/s41579-019-0222-5 Comasio, A., Verce, M., Van Kerrebroeck, S., & De Vuyst, L. (2020). Diverse Microbial Composition of Sourdoughs From Different Origins. Frontiers in Microbiology , 11 . https://www.frontiersin.org/journals/microbiology/articles/10.3389/fmicb.2020.01212 Constant, N. L., & Hughes, J. (2023). Diversifying citizen science through the inclusion of young people. Journal of Science Communication , 22 (2), A07. https://doi.org/10.22323/2.22020207 Courtin, C. M., & Delcour, J. A. (2002). Arabinoxylans and Endoxylanases in Wheat Flour Bread-making. Journal of Cereal Science , 35 (3), 225–243. https://doi.org/10.1006/jcrs.2001.0433 D’Amico, V., Gänzle, M., Call, L., Zwirzitz, B., Grausgruber, H., D’Amico, S., & Brouns, F. (2023). Does sourdough bread provide clinically relevant health benefits? Frontiers in Nutrition , Volume 10-2023 . https://www.frontiersin.org/journals/nutrition/articles/10.3389/fnut.2023.1230043 De Vuyst, L., Comasio ,Andrea, & and Kerrebroeck, S. V. (2023). Sourdough production: Fermentation strategies, microbial ecology, and use of non-flour ingredients. Critical Reviews in Food Science and Nutrition , 63 (15), 2447–2479. https://doi.org/10.1080/10408398.2021.1976100 De Vuyst, L., Van Kerrebroeck, S., & Leroy, F. (2017). Chapter Two—Microbial Ecology and Process Technology of Sourdough Fermentation. In S. Sariaslani & G. M. Gadd (Eds.), Advances in Applied Microbiology (Vol. 100, pp. 49–160). Academic Press. https://doi.org/10.1016/bs.aambs.2017.02.003 Deleu, L. J., Lemmens, E., Redant, L., & Delcour, J. A. (2020). The major constituents of rye (Secale cereale L.) flour and their role in the production of rye bread, a food product to which a multitude of health aspects are ascribed. Cereal Chemistry , 97 (4), 739–754. https://doi.org/10.1002/cche.10306 Fleming, E. E., Ziegler, G. R., & Hayes, J. E. (2015). Check-all-that-apply (CATA), sorting, and polarized sensory positioning (PSP) with astringent stimuli. Food Quality and Preference , 45 , 41–49. https://doi.org/10.1016/j.foodqual.2015.05.004 Fraisl, D., Hager, G., Bedessem, B., Gold, M., Hsing, P.-Y., Danielsen, F., Hitchcock, C. B., Hulbert, J. M., Piera, J., Spiers, H., Thiel, M., & Haklay, M. (2022). Citizen science in environmental and ecological sciences. Nature Reviews Methods Primers , 2 (1), 1–20. https://doi.org/10.1038/s43586-022-00144-4 Garcia, S., Ordoñez, S., López-Molina, V. M., Lacruz-Pleguezuelos, B., Carrillo de Santa Pau, E., & Marcos-Zambrano, L. J. (2023). Citizen science helps to raise awareness about gut microbiome health in people at risk of developing non-communicable diseases. Gut Microbes , 15 (1), 2241207. https://doi.org/10.1080/19490976.2023.2241207 Gupta, A., Gupta, R., & Singh, R. L. (2016). Microbes and Environment. Principles and Applications of Environmental Biotechnology for a Sustainable Future , 43. https://doi.org/10.1007/978-981-10-1866-4_3 Hernández-Velázquez, R., Flörl, L., Lavrinienko, A., Sebechlebská, Z., Merk, L., Greppi, A., & Bokulich, N. A. (2024). The future is fermented: Microbial biodiversity of fermented foods is a critical resource for food innovation and human health. Trends in Food Science & Technology , 150 , 104569. https://doi.org/10.1016/j.tifs.2024.104569 Landis, E. A., Oliverio, A. M., McKenney, E. A., Nichols, L. M., Kfoury, N., Biango-Daniels, M., Shell, L. K., Madden, A. A., Shapiro, L., Sakunala, S., Drake, K., Robbat, A., Booker, M., Dunn, R. R., Fierer, N., & Wolfe, B. E. (2021). The diversity and function of sourdough starter microbiomes. ELife , 10 , e61644. https://doi.org/10.7554/eLife.61644 Lewandowski, E., Caldwell, W., Elmquist, D., & Oberhauser, K. (2017). Public Perceptions of Citizen Science. Citizen Science: Theory and Practice , 2 (1). https://doi.org/10.5334/cstp.77 Marco, M. L., Sanders, M. E., Gänzle, M., Arrieta, M. C., Cotter, P. D., De Vuyst, L., Hill, C., Holzapfel, W., Lebeer, S., Merenstein, D., Reid, G., Wolfe, B. E., & Hutkins, R. (2021). The International Scientific Association for Probiotics and Prebiotics (ISAPP) consensus statement on fermented foods. Nature Reviews Gastroenterology & Hepatology , 18 (3), 196–208. https://doi.org/10.1038/s41575-020-00390-5 Pateman, R. M., de Bruin, A., Piirsalu, E., Reynolds, C., Stokeld, E., & West, S. E. (2020). Citizen Science for Quantifying and Reducing Food Loss and Food Waste. Frontiers in Sustainable Food Systems , 4 . https://doi.org/10.3389/fsufs.2020.589089 Piochi, M., Cabrino, G., & Torri, L. (2021). Check-All-That-Apply (CATA) Test to Investigate the Consumers’ Perception of Olive Oil Sensory Properties: Effect of Storage Time and Packaging Material. Foods , 10 (7), 1551. https://doi.org/10.3390/foods10071551 Reese, A. T., Madden, A. A., Joossens, M., Lacaze, G., & Dunn, R. R. (2020). Influences of Ingredients and Bakers on the Bacteria and Fungi in Sourdough Starters and Bread. MSphere , 5 (1), 10.1128/msphere.00950-19. https://doi.org/10.1128/msphere.00950-19 Ribet, L., Dessalles, R., Lesens, C., Brusselaers, N., & Durand-Dubief, M. (2023). Nutritional benefits of sourdoughs: A systematic review. Advances in Nutrition , 14 (1), 22–29. https://doi.org/10.1016/j.advnut.2022.10.003 Rizzello, C. G., Portincasa, P., Montemurro, M., Di Palo, D. M., Lorusso, M. P., De Angelis, M., Bonfrate, L., Genot, B., & Gobbetti, M. (2019). Sourdough Fermented Breads are More Digestible than Those Started with Baker’s Yeast Alone: An In Vivo Challenge Dissecting Distinct Gastrointestinal Responses. Nutrients , 11 (12), 2954. https://doi.org/10.3390/nu11122954 Valentino, V., Magliulo, R., Farsi, D., Cotter, P. D., O’Sullivan, O., Ercolini, D., & De Filippis, F. (2024). Fermented foods, their microbiome and its potential in boosting human health. Microbial Biotechnology , 17 (2), e14428. https://doi.org/10.1111/1751-7915.14428 Walsh, L. H., Breselge, S., Martin, J. G. P., Coakley, M., Ferguson, E., Stapleton, A., Crispie, F., O’Toole, P. W., & Cotter, P. D. (2024). Kefir4All, a citizen science initiative to raise awareness of the roles that microbes play in food fermentation. Journal of Microbiology & Biology Education , 0 (0), e00155-23. https://doi.org/10.1128/jmbe.00155-23 Waugh, J. K., Lindsey, J. K., Stewart, M. Z., Winter, J. C., & Parrish, J. K. (2023). Demographics of Public Participation in Science: A Meta-Analytic Approach. Citizen Science: Theory and Practice , 8 (1). https://doi.org/10.5334/cstp.610 Weigelhofer, G., & Pölz, E.-M. (2016). Data Quality in Citizen Science Projects: Challenges and Solutions. Frontiers in Environmental Science , 4 . https://doi.org/10.3389/conf.FENVS.2016.01.00011 Wiggins, A., & Wilbanks, J. (2019). The Rise of Citizen Science in Health and Biomedical Research. The American Journal of Bioethics , 19 (8), 3–14. https://doi.org/10.1080/15265161.2019.1619859 Additional Declarations No competing interests reported. Supplementary Files Supplementarydata.pdf Cite Share Download PDF Status: Published Journal Publication published 27 Jan, 2026 Read the published version in npj Science of Food → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6627352","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":464259073,"identity":"b6d01192-f4df-4fc4-9f03-086b28921ffa","order_by":0,"name":"Annina Meyer","email":"","orcid":"","institution":"ETH Zurich","correspondingAuthor":false,"prefix":"","firstName":"Annina","middleName":"","lastName":"Meyer","suffix":""},{"id":464259074,"identity":"57b5b4d6-d985-4b24-bd14-c19c084c0a59","order_by":1,"name":"Thomas Gettemans","email":"","orcid":"","institution":"Vrije Universiteit Brussel (VUB)","correspondingAuthor":false,"prefix":"","firstName":"Thomas","middleName":"","lastName":"Gettemans","suffix":""},{"id":464259075,"identity":"f7b72ce9-9299-4067-b50b-ae5005ce3c12","order_by":2,"name":"Jan Patrick Tan","email":"","orcid":"","institution":"ETH Zurich","correspondingAuthor":false,"prefix":"","firstName":"Jan","middleName":"Patrick","lastName":"Tan","suffix":""},{"id":464259076,"identity":"a11cda35-aa36-4248-be3d-e06c90ecce48","order_by":3,"name":"Fabio Tuccillo","email":"","orcid":"","institution":"University of Helsinki","correspondingAuthor":false,"prefix":"","firstName":"Fabio","middleName":"","lastName":"Tuccillo","suffix":""},{"id":464259077,"identity":"12734e76-e0b2-40b7-8828-277a0481e34b","order_by":4,"name":"Chiara Viretto","email":"","orcid":"","institution":"Free University of Bolzano-Bozen","correspondingAuthor":false,"prefix":"","firstName":"Chiara","middleName":"","lastName":"Viretto","suffix":""},{"id":464259078,"identity":"cfb8e220-d047-4a7f-9aee-0083ccbe41ce","order_by":5,"name":"Iulia-Roxana Angelescu","email":"","orcid":"","institution":"Institute of Biology Bucharest of the Romanian Academy","correspondingAuthor":false,"prefix":"","firstName":"Iulia-Roxana","middleName":"","lastName":"Angelescu","suffix":""},{"id":464259079,"identity":"b9cf6eb6-a44f-4f23-ad70-3d00463186db","order_by":6,"name":"Yamina Bondt","email":"","orcid":"","institution":"KU Leuven","correspondingAuthor":false,"prefix":"","firstName":"Yamina","middleName":"","lastName":"Bondt","suffix":""},{"id":464259080,"identity":"35550055-7d8a-4b49-8d6a-d7fe4527d26f","order_by":7,"name":"Michelle Neugebauer","email":"","orcid":"","institution":"ETH Zurich","correspondingAuthor":false,"prefix":"","firstName":"Michelle","middleName":"","lastName":"Neugebauer","suffix":""},{"id":464259081,"identity":"7a3de913-1d7e-45ee-8fe2-78426f82919f","order_by":8,"name":"Ali Zein Alabiden Tlais","email":"","orcid":"","institution":"Free University of Bolzano-Bozen","correspondingAuthor":false,"prefix":"","firstName":"Ali","middleName":"Zein Alabiden","lastName":"Tlais","suffix":""},{"id":464259082,"identity":"b448cd8d-0d59-4cf3-b43c-be319e8e27fb","order_by":9,"name":"Fabio Cavelti","email":"","orcid":"","institution":"ETH Zurich","correspondingAuthor":false,"prefix":"","firstName":"Fabio","middleName":"","lastName":"Cavelti","suffix":""},{"id":464259083,"identity":"da9c2f2c-4b0e-4710-a5f2-9681dfb9f186","order_by":10,"name":"Luc Vuyst","email":"","orcid":"","institution":"Vrije Universiteit Brussel (VUB)","correspondingAuthor":false,"prefix":"","firstName":"Luc","middleName":"","lastName":"Vuyst","suffix":""},{"id":464259084,"identity":"4991b3dc-81af-4d84-9104-a83443e752ad","order_by":11,"name":"Marco Gobbetti","email":"","orcid":"","institution":"Free University of Bolzano-Bozen","correspondingAuthor":false,"prefix":"","firstName":"Marco","middleName":"","lastName":"Gobbetti","suffix":""},{"id":464259085,"identity":"0218f6b8-81fe-49c2-87e9-19eff5cc4f24","order_by":12,"name":"Christophe M. Courtin","email":"","orcid":"","institution":"KU Leuven","correspondingAuthor":false,"prefix":"","firstName":"Christophe","middleName":"M.","lastName":"Courtin","suffix":""},{"id":464259086,"identity":"cd944d2e-eb45-48dc-8487-b1b9aafc28d9","order_by":13,"name":"Medana Zamfir","email":"","orcid":"","institution":"Institute of Biology Bucharest of the Romanian Academy","correspondingAuthor":false,"prefix":"","firstName":"Medana","middleName":"","lastName":"Zamfir","suffix":""},{"id":464259087,"identity":"c7c7d410-d08d-4467-9f6a-f631401acc9b","order_by":14,"name":"Rossana Coda","email":"","orcid":"","institution":"University of Helsinki","correspondingAuthor":false,"prefix":"","firstName":"Rossana","middleName":"","lastName":"Coda","suffix":""},{"id":464259088,"identity":"3e68b166-c6e8-4570-ab19-8ca9e2e58f7a","order_by":15,"name":"Laura Nyström","email":"","orcid":"","institution":"ETH Zurich","correspondingAuthor":false,"prefix":"","firstName":"Laura","middleName":"","lastName":"Nyström","suffix":""},{"id":464259089,"identity":"b92cbbd2-4d60-4273-9733-7750ca1e089e","order_by":16,"name":"Stefan Weckx","email":"","orcid":"","institution":"Vrije Universiteit Brussel (VUB)","correspondingAuthor":false,"prefix":"","firstName":"Stefan","middleName":"","lastName":"Weckx","suffix":""},{"id":464259090,"identity":"e06d4999-9278-43e1-88a1-0f430804682a","order_by":17,"name":"Nicholas A. Bokulich","email":"data:image/png;base64,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","orcid":"","institution":"ETH Zurich","correspondingAuthor":true,"prefix":"","firstName":"Nicholas","middleName":"A.","lastName":"Bokulich","suffix":""}],"badges":[],"createdAt":"2025-05-09 09:38:24","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6627352/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6627352/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41538-026-00725-7","type":"published","date":"2026-01-27T15:58:41+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":83883282,"identity":"5ededed1-4045-4713-9a4a-4e828c44a256","added_by":"auto","created_at":"2025-06-04 06:09:07","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":227061,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCo-design-based study setup for conducting sourdough fermentation related surveys and sampling sourdoughs across Europe.\u003c/strong\u003e Study workflow: Co-design workshops and protocol harmonization (1) were followed by project advertising (2) and online recruitment (Survey 1) (3), local screening, and distribution of sampling kits (4). Participants conducted home experiments (sourdough/bread pH, bread density, sensory evaluation), reported results via Survey 2 (5), and mailed sourdough samples to hub labs. Samples were processed, biobanked (6), and metadata (resources 1, 2, and scientist-recorded data) curated and translated for integration and analysis (7). Icons used in this figure were sourced from Canva (\u003ca href=\"http://www.canva.com/\" target=\"_new\"\u003ewww.canva.com\u003c/a\u003e) and edited in Microsoft PowerPoint.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6627352/v1/55dfbca074d8b39df03251aa.png"},{"id":83883284,"identity":"49c66f18-f956-44cb-8060-f37c308a88c0","added_by":"auto","created_at":"2025-06-04 06:09:07","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":441070,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSourdough citizen science participation and participant profiles.\u003c/strong\u003e \u003cstrong\u003ea,\u003c/strong\u003e Global distribution of registrations and collected sourdough samples, with highest coverage in Europe. Overlapping dots indicate densely sampled regions. \u003cstrong\u003eb, \u003c/strong\u003eRegistrations and sample collections over time, from project launch to campaign closure. \u003cstrong\u003ec,\u003c/strong\u003e Distribution of sign-ups and collected samples across top-contributing countries. Countries with fewer than five registered sourdoughs are grouped under “Other.” Hub countries show the highest coverage. \u003cstrong\u003ed,\u003c/strong\u003eLanguage distribution of all received registrations (n = 1131), including language splits per CS hub. Segment labels indicate absolute counts. \u003cstrong\u003ee, \u003c/strong\u003eMost effective advertising channels by country. Associations assessed via Chi-squared tests; only significant associations shown. \u003cstrong\u003ef,\u003c/strong\u003e Distribution of sourdough sources. “Bakery” includes all non-household origins (e.g., artisanal bakeries, industry, restaurants). Samples with unknown origins excluded (n = 3). \u003cstrong\u003eg, \u003c/strong\u003eReported origins of registered sourdoughs. \u003cstrong\u003eh,\u003c/strong\u003eParticipant experience with sourdough, ranging from less than half to more than 5 years, including professional bakers. \u003cstrong\u003ei,\u003c/strong\u003eYears when sourdoughs started.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-6627352/v1/0f37014caa3981c0890da70b.png"},{"id":83882634,"identity":"f2d7a279-a221-4fac-a17c-87e67ac2dd0d","added_by":"auto","created_at":"2025-06-04 06:01:07","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":219103,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eVariations in the registered process and environmental parameters, ingredients and fermentation vessels used in bakeries (B; 109 responses) and households (H; 1019 responses).\u003c/strong\u003e \u003cstrong\u003ea, \u003c/strong\u003eDistribution of fermentation related parameters: the backslopping frequency, the location of the sourdough during fermentation, the temperature and time during fermentation and the temperature of storage of the sourdough between feeding moments. \u003cstrong\u003eb, \u003c/strong\u003eDistribution of various environmental parameters: sourdough age (categories defined with the lower limit included and the upper limit excluded), origin of the sourdough, the skill level of the sourdough baker, the ownership of pets and the presence of kids and plants. \u003cstrong\u003ec,\u003c/strong\u003e Distribution of sourdough ingredient related parameters: Dough yield (categories defined with the lower limit included and the upper limit excluded), flour type (\u003cem\u003eTriticum aestivum\u003c/em\u003e is referred to as \"wheat,\" \u003cem\u003eTriticum spelta\u003c/em\u003e as \"spelt,\" and \u003cem\u003eSecale cereale\u003c/em\u003eas \"rye\"), the use of organic flour, the prior use of bakers yeast in the environment where the sourdough is maintained, and the water type used when feeding the sourdough. \u003cstrong\u003ed,\u003c/strong\u003eCharacteristics related to the vessel in which the sourdough is maintained.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-6627352/v1/e3d1caf813f40e798d842964.png"},{"id":83883703,"identity":"778a0d3f-1b0f-40f2-bb99-5b3cd37c0a55","added_by":"auto","created_at":"2025-06-04 06:17:07","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":211586,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePersonal motivations and perceptions related to sourdough baking.\u003c/strong\u003e \u003cstrong\u003ea, \u003c/strong\u003eParticipant-reported motivations for baking with sourdough, assessed using a \"check-all-that-apply\" format. \u003cstrong\u003eb,\u003c/strong\u003e Perceived health benefits of sourdough bread baking and consumption. Free-text responses were thematically categorized into nine main benefit categories. \u003cstrong\u003ec,\u003c/strong\u003e Chi-square analyses of associations between selected baking motivations and perceived health benefits and the categories flour milling grade, grain type, and use of organic versus non-organic flour. Positive and negative associations are color-coded; bubble size reflects effect size (Cramér’s V), and statistically significant associations are highlighted. Associations are hierarchically clustered per category (flour milling grade and grain type) based on directional Cramér’s V.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-6627352/v1/0b3dade18d56129cce7926cd.png"},{"id":83882641,"identity":"dc0214b3-0b6a-4dbb-823a-f454b6754afb","added_by":"auto","created_at":"2025-06-04 06:01:07","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":510408,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCitizen science contributions to sourdough characterization and sensory profiling. a,\u003c/strong\u003e Correlation between sourdough and bread pH values, grouped by flour type used for feeding the sourdoughs; bubble size reflects bread density; significant Pearson correlations indicated (*p\u0026lt;0.05, **p\u0026lt;0.01, ***p\u0026lt;0.001). \u003cstrong\u003eb,\u003c/strong\u003e Comparison of pH measured by citizens vs. laboratory measurements post-shipping under non-temperature-controlled conditions. \u003cstrong\u003ec,\u003c/strong\u003eReported sourdough odours ranked by frequency and coloured by flavour category (n = 557); average occurrence indicates mean frequency of sensory notes. \u003cstrong\u003ed,\u003c/strong\u003e Principal Coordinates Analysis (PCoA) of Jaccard dissimilarity matrix for odours; top 10 aroma features contributing most to variance shown, coloured by sourdough flour type. \u003cstrong\u003e(e, f)\u003c/strong\u003e Pearson correlations of aroma categories with feeding parameters and experimental measurements \u003cstrong\u003e(e),\u003c/strong\u003eand sourdough flour types \u003cstrong\u003e(f)\u003c/strong\u003e; features and predictors hierarchically clustered by correlation direction.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-6627352/v1/fec72fdb1a96673d38ee0c9d.png"},{"id":83882638,"identity":"a9693de8-29f7-4078-86ad-1d731c1f5fab","added_by":"auto","created_at":"2025-06-04 06:01:07","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":201071,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSubstrate-dependent variation in sourdough physicochemical properties and maintenance practices.\u003c/strong\u003e \u003cstrong\u003ea,\u003c/strong\u003e Comparison of dough yield between wheat- and rye-based sourdoughs. \u003cstrong\u003eb, \u003c/strong\u003eParticipant-reported pH values using pH strips. \u003cstrong\u003ec, \u003c/strong\u003eTotal titratable acidity (TTA) measured in the laboratory, defined as the volume (mL) of 0.1 M NaOH required to reach pH 8.5 in 10 g of sourdough diluted with 100 g deionized water. \u003cstrong\u003ed,\u003c/strong\u003e Bread density of breads made with wheat or rye sourdough, as reported by participants. Values \u0026gt;1 were considered erroneous and excluded from analysis. \u003cstrong\u003e(a;b;c;d)\u003c/strong\u003e Statistical comparisons were performed using the Mann–Whitney U test with BH-FDR correction (p ≥ 0.05 = ns; * \u0026lt; 0.05; ** \u0026lt; 0.01; *** \u0026lt; 0.001). \u003cstrong\u003ee,\u003c/strong\u003e Hierarchical clustering of sourdoughs by substrate, based on Pearson correlations between flour compositions (including mixed flours), feeding parameters, and sourdough characteristics. Bubble size reflects absolute correlation strength (|r|), and color indicates direction (positive/negative).\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-6627352/v1/a91bde189501981262075da6.png"},{"id":83882635,"identity":"f360cd11-d361-4966-a71c-08fe9dcc8176","added_by":"auto","created_at":"2025-06-04 06:01:07","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":545411,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eRegional variation in sourdough ingredients, maintenance practices, and physicochemical characteristics among European citizen scientists.\u003c/strong\u003e \u003cstrong\u003ea,\u003c/strong\u003e Bubble plot of significant Chi-square associations showing regional differences in ingredients and maintenance practices (e.g., non-flour additions, flour type changes, milling grade, organic use, grain base, water source, baker’s yeast). \u003cstrong\u003eb,\u003c/strong\u003e Heatmap of baking motivations and perceived health benefits by country and self-reported skill level. \u003cstrong\u003e(a,b)\u003c/strong\u003e Only categories with at least one significant association are shown. Associations are color-coded by enrichment direction and significance (BH-FDR p \u0026lt; 0.05: dark; uncorrected p \u0026lt; 0.05: light); bubble size reflects effect size (Cramér’s V). Countries are clustered by Cramér’s V profiles. \u003cstrong\u003ec,\u003c/strong\u003e PCA of numeric metadata: feeding parameters, home experiment results, TTA, and latitude (excluding substrate). \u003cstrong\u003ed,\u003c/strong\u003ePCA including feeding parameters, home results, lab TTA, and flour proportions (excluding geography). \u003cstrong\u003e(c,d)\u003c/strong\u003e Samples with missing values were excluded. Points are colored by flour type or country; ellipses show 95% confidence intervals for household vs. bakery sources. Groups with \u0026lt;5 samples were excluded.\u003c/p\u003e","description":"","filename":"7.png","url":"https://assets-eu.researchsquare.com/files/rs-6627352/v1/12202be7bd90ccea33036d96.png"},{"id":101690581,"identity":"db10bdc7-75bd-4857-9dea-9adda312d64f","added_by":"auto","created_at":"2026-02-02 16:06:00","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3594813,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6627352/v1/a5a0e7e5-f9cd-4934-a100-34da4ee30f03.pdf"},{"id":83883283,"identity":"d7c4d000-1c1e-4da2-898b-499a88ee970f","added_by":"auto","created_at":"2025-06-04 06:09:07","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":910915,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementarydata.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6627352/v1/a331ab55df0e03e964117df5.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Rising together: Exploring Sourdough Fermentation Diversity through Co-design in the HealthFerm Citizen Science Initiative","fulltext":[{"header":"Introduction","content":"\u003cp\u003eFermented foods, produced through desired microbial growth and enzymatic activities (Marco et al., \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), are the primary natural source of microorganisms in the human diet, and hold significant cultural, nutritional, and ecological values (Hern\u0026aacute;ndez-Vel\u0026aacute;zquez et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Fermented foods are popularly perceived as beneficial for health, possibly due to the nature of fermentation as a traditional, low-input food processing technique, and source of dietary microorganisms. However, despite their roles in shaping fermented food products, supporting regional biodiversity, influencing gut health (Valentino et al., \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), and even contributing to planetary well-being (Cavicchioli et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Gupta et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), the functional significance of microorganisms in food production and health remains largely overlooked by the general public.\u003c/p\u003e \u003cp\u003eStudying food microbiomes presents significant hurdles, particularly in obtaining samples that reflect diverse geographical, cultural, and procedural origins, a limitation compounded by the constrained reach of traditional research approaches. Citizen science (CS) offers a transformative solution to overcome these challenges while also strengthening the connection between scientists and the public. Through public involvement in microbiome research \u0026mdash; ranging from simple contributory efforts to fully co-designed initiatives \u0026mdash; scientists can gather more representative datasets that reflect the ecological, cultural, culinary, or health-related diversity of microbiomes. Beyond expanding the scope of research, CS fosters community participation in scientific discovery, mutual learning and creates awareness of the critical roles microorganisms play in daily life (Garcia et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). It also acts as a platform for innovation, public engagement, and evidence-based policy advocacy, ensuring the microbial world\u0026rsquo;s significance is recognised and valued across disciplines and societies (Fraisl et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIntegrating CS into system-level microbiome studies unlocks insights essential for addressing global challenges such as antimicrobial resistance, climate change, and food security (Lewandowski et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Pateman et al., \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). In fermented food research, it also highlights overlooked microbial ecosystems and deepens public and academic understanding of microbial diversity, bridges scientific research with traditional knowledge, and underscores its role in sustainable food systems, cultural heritage, and human and planetary health. For example, Kefir4All and the Global Sourdough Project highlighted the roles of microbial biodiversity in shaping the production and quality of home-made fermented foods (Landis et al., \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Walsh et al., \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Through extensive public participation, diverse outreach, and transparent dissemination of findings, these projects showcase the power of CS to generate robust datasets, challenge prevailing assumptions, and strengthen connections between scientists and the public.\u003c/p\u003e \u003cp\u003eSourdough, a traditional fermented mixture of water and cereal flour(s) with roots traceable to ancient times, has played a key role in the evolution of diverse food cultures by enhancing the sensory attributes, texture, nutritional value, and microbiological safety of cereal-based foods (Cappelle et al., \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). Fermentation allows microbial growth and metabolism in the dough, producing sourdough that can serve as a leavening agent and can be propagated through backslopping\u0026mdash;adding a portion of fermented dough to fresh flour and water (Arora et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Calvert et al., \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; De Vuyst et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2017\u003c/span\u003e, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Despite its simple ingredients, traditional sourdoughs exhibit remarkable microbial variability, influenced by differences in household and bakery practices, flour type, dough yield, backslopping ratio, and fermentation/storage conditions (Comasio et al., \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; De Vuyst et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Landis et al., \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). As most studies to date focus on laboratory-prepared or small-scale bakery sourdoughs, comprehensive large-scale investigations of spontaneously fermented (Type 1) sourdoughs in real-world settings are still needed (Bondt et al., \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eHere, we describe the design and execution of a large-scale sourdough fermentation study, utilizing a CS co-design approach to facilitate greater collaboration with participants. The study was conducted within the framework of the HealthFerm project (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://healthferm.eu/\u003c/span\u003e\u003cspan address=\"https://healthferm.eu/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e)\u003c/span\u003e, a European initiative exploring innovative pulse- and cereal-based food fermentations, their potential health effects, and consumer perceptions. Through its citizen science component, HealthFerm aims to advance understanding of the sourdough ecosystem and develop novel microbial resources to support the targeted redesign of fermented foods, promoting healthier and more sustainable diets. Here, we highlight important study design considerations as well as the technical challenges and potential biases inherent to this approach and present sourdough citizen science insights. Key outcomes include broad participant engagement, the creation of geographically diverse datasets, and valuable insights for future CS initiatives. Altogether, the project explored sourdough maintenance practices and consumer preferences and perceptions across Europe, establishing a framework for microbiologists designing collaborative, volunteer-driven microbiome research.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eCollaborative, iterative co-design guides methodological citizen science implementation and refinement\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eCollaborative methodological planning was carried out with HealthFerm partner institutions and through direct engagement with citizens. Five universities served as local citizen science centers (hubs) to promote the study, recruit participants, and collect sourdough samples (Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003ea). Three hubs organized co-design workshops located in Herent (Belgium), Helsinki (Finland), and Zurich (Switzerland), involving 70 participants (27, 17, and 26, respectively). These workshops played a central role in shaping and refining methodologies throughout the project (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003eb). Participants showed strong engagement and willingness to conduct at-home experiments and complete follow-up questionnaires. Their feedback informed improvements to the sampling kits, such as simplifying terminology, adding a project overview, and clarifying return instructions. Participant questions and interests were gathered to guide dissemination across project stages (Supplementary list 1).\u003c/p\u003e \u003cp\u003eParticipants from different countries reported varying practices for sourdough refreshment, particularly in recipe precision. In Helsinki, only 66% used scales and exact recipes, compared to 85% in Zurich, reflecting a preference for volumetric units in Finland. Therefore, to standardize at-home measurements, a volume-to-mass conversion table was included in the instruction booklet. Feedback also guided decisions on sample submission and communication preferences. Local postal services were selected as the main delivery method, with optional drop-off for those near hub labs; sampling kits were distributed exclusively by mail. In response to requests for centralized updates, a multilingual blog was added to the HealthFerm website (Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003eb).\u003c/p\u003e \u003cp\u003eEngagement patterns and sample contributions in a trans-European citizen science study\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eExtensive outreach across Europe led to strong engagement primarily from European as well as some non-European participants (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea). By September 2024, the project had registered 1091 sourdoughs and collected 671 samples from 591 participants across 33 countries, including 27 in Europe (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb). Most sign-ups (68.6%) and samples (65.7%) originated from hub countries (Switzerland, Finland, Belgium, Italy, Romania) and Germany (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec), resulting in a slightly skewed language distribution (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed). 61 individuals submitted more than one sourdough (n\u0026thinsp;=\u0026thinsp;2\u0026ndash;5).\u003c/p\u003e \u003cp\u003eRecruitment success varied by country and channel (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ee, Fig. S2a): sourdough blogs were particularly effective in Germany, press in Switzerland, LinkedIn in Denmark, and Facebook in Finland (Table S2). In contrast, word-of-mouth in Finland and Facebook in Switzerland and Germany were less impactful.\u003c/p\u003e \u003cp\u003eMerging sign-up forms and survey responses showed that 87.2% of sourdoughs came from household bakers (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ef), with 67.7% self-initiated and 8.0% passed on by others (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eg). Regarding experience, 26.6% had baked for under two years, while over a third had five or more years of experience or were professionals (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eh). Most sourdoughs were less than a year old at sampling (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ei), with a notable peak of newly started sourdoughs in 2020, reflecting the surge of popularity in sourdough baking during the COVID-19 pandemic; the oldest registered sourdough dated back to 1865.\u003c/p\u003e \u003cp\u003eFermentation routines and storage conditions differ between households and bakeries\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eDifferences in environmental and fermentation practices were observed between professional (n\u0026thinsp;=\u0026thinsp;109) and home bakers (n\u0026thinsp;=\u0026thinsp;1019) (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Sourdoughs from both groups were proportionally distributed across European countries, minimizing regional bias (Fig. S2b). Professional bakers more often maintained older starters (55% \u0026gt; 5 years vs. 28% in households), refreshed them daily (41.3% vs. 43.7% weekly in households), and fermented at higher temperatures (23\u0026deg;C vs. 21\u0026deg;C) with shorter fermentation times. Both groups typically stored sourdoughs at or below 10\u0026deg;C when not in use. Plastic containers were more common in bakeries (61.5%), while home bakers preferred glass jars (only 9.4% use plastic containers). The presence of plants, pets, or children in the baking environment did not differ significantly, likely due to unclear survey wording. Ingredients were similar across groups (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec): wheat was most common (52.3% bakeries; 53.4% households), followed by rye, with widespread use of organic flours (66.1% bakeries; 59.6% households). Most reported dough yields (DY; \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\frac{Flour\\:\\left(g\\right)\\:+\\:water\\:\\left(g\\right)}{Flour\\:\\left(g\\right)}*100\\)\u003c/span\u003e\u003c/span\u003e) between 150 and 250, indicating firm to semi-liquid sourdoughs.\u003c/p\u003e \u003cp\u003eBaking motivations and perceived health benefits shape ingredient choices\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e Personal motivations and beliefs about sourdough varied between participants. Participants most frequently cited hedonistic motivations such as taste and a general interest in baking, and financial/tradition motivations less frequently (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea). A majority (84.2%) perceived sourdough as healthier than yeast-leavened bread, with no significant difference across experience levels (Fig. S3a, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.115). Gastrointestinal benefits were the most reported perceived health benefit, again consistent across levels of baking experience (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb, Fig. S3b, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.115). Health-motivated bakers significantly more frequently used rye and less frequently wheat as the grain base (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec, Table S3). Tradition-driven bakers more often used rye and spelt, while rye was also overrepresented among those citing health-related reasons such as using \"healthier ingredients\" or experiencing \"fewer intolerances\" compared to baking with conventional yeast (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec, Table S3). Bakers using wholemeal or flour mixes more often reported enjoyment, health motivation, and gastrointestinal benefits, compared to those using endosperm or unspecified flour types, who were less likely to cite enjoyment, taste, or digestibility (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec, Table S3). Health-motivated bakers also used organic flour significantly more often and associated it with improved nutritional quality or bioavailability (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec, Table S3).\u003c/p\u003e \u003cp\u003eCitizen science experiments reveal patterns and challenges in sourdough physicochemical and sensory characteristics\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eAmong participants who submitted sourdough samples, 63.6% (427/671) completed five optional home experiments, while 15.6% did not perform any. In 13.8% of cases, no sourdough bread was baked, and 8.6% of those who baked did not complete the experiments. pH measurements performed at home revealed a positive correlation between sourdough and bread pH (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea), indicating that sourdough acidification generally yielded more acidic bread, regardless of flour type. Bread density, however, was not affected by pH (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea, Fig. S4). Laboratory-measured pH values were on average 0.73 units lower than home measurements (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eb, Fig. S5), likely due to further acidification during shipping and/or the use of accurate laboratory pH meters versus pH strips at home.\u003c/p\u003e \u003cp\u003e Participants characterised the aroma of their sourdoughs using the Check-All-That-Apply (CATA) methodology and selected from 56 predefined descriptors, organised into 13 categories (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ec). Fermented, grain/cereal, fermented dairy, fruity, and sour aromas were most prevalent; ocean, animal, and body odour were least common. High variability in aroma perception was evident in Jaccard dissimilarity-based PCoA (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ed), where the first two components explained only 13.7% of the total variance. Features contributing most to the variance included yeasty, bread, and yoghurt all positively correlated with PC1, whereas porridge, hay, lactic acid, alcoholic, vinegar and wholegrain aromas also ranked in the top ten features but negatively correlated with PC1 (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ed). PERMANOVA identified categorical variables such as plant presence, organic flour use, milling grade, and acidic beverage fermentation as significantly associated with aroma variation (BH-FDR p\u0026thinsp;\u0026lt;\u0026thinsp;0.05), although effect sizes were low (R\u0026sup2; \u0026lt; 0.005; Table S4). Among numerical variables, wheat proportion and the number of maintained sourdoughs (positive correlation) and backslopping frequency (negative correlation), were significant but weak predictors (Mantel Pearson R\u0026thinsp;\u0026lt;\u0026thinsp;0.1; Table S5). Notably, not clear separation was observed when grouping samples by flour type in the full aroma feature space (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ed, Table S5).\u003c/p\u003e \u003cp\u003eNonetheless, specific aroma descriptors in isolation correlated with experimental data and flour types. Sourdoughs described as \u0026ldquo;sour\u0026rdquo; had significantly higher TTAs (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ef, Table S6) and were more frequently composed of wholemeal rye flour (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eg). Wheat sourdoughs demonstrated distinct aroma characteristics (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eg), were less sour and more often associated with chemical odours, which negatively correlated with TTA (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ef, Table S6).\u003c/p\u003e \u003cp\u003eSubstrate-dependent variation in sourdough composition, maintenance, and physicochemical properties\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe cereal used for sourdough influences its physicochemical properties and associated recipes. Wheat and rye, the most commonly used cereals for sourdough feeding (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec), were selected for in-depth analysis. Reflecting rye\u0026rsquo;s higher water absorption, participants using rye maintained sourdoughs at higher dough yield than those using wheat (r\u0026thinsp;=\u0026thinsp;0.30, p\u0026thinsp;=\u0026thinsp;3.10\u0026times;10⁻\u0026sup1;⁸; Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ea). Rye sourdoughs also had slightly lower participant-measured pH (r\u0026thinsp;=\u0026thinsp;0.12, p\u0026thinsp;=\u0026thinsp;8.84\u0026times;10⁻⁴) and higher TTA values (r\u0026thinsp;=\u0026thinsp;0.37, p\u0026thinsp;=\u0026thinsp;3.0\u0026times;10⁻\u0026sup2;⁶), despite similar lab pH (p\u0026thinsp;=\u0026thinsp;0.15) (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eb,c). Participants baking with rye reported denser breads than those using wheat (r\u0026thinsp;=\u0026thinsp;0.37, p\u0026thinsp;=\u0026thinsp;3.48\u0026times;10⁻\u0026sup1;\u0026sup1;; Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ed). These differences were mirrored in sourdough clustering based on Pearson correlations between proportion of flour type used per sourdough and physicochemical or feeding parameters (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ee). Sourdoughs with high rye content (wholemeal or not) formed a distinct cluster, associated with higher TTA and dough yield, and lower pH and backslopping frequency. Sourdoughs with non-rye wholemeal flours or high spelt content showed intermediate profiles between rye- and wheat-based sourdoughs (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ee; Table S7).\u003c/p\u003e \u003cp\u003eFermentation fingerprints: how region, substrate, and role define sourdough fermentation practices\u003c/p\u003e \u003cp\u003eSourdough ingredient choices varied regionally across Europe, shaped by cultural traditions, climate, and lifestyle. Chi-square enrichment analysis revealed significant country-specific differences in flour types, milling grades, water sources, added ingredients, flour switching, organic use, and concurrent yeast usage (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ea, Table S8). Clustering countries by directional Cram\u0026eacute;r\u0026rsquo;s V effect sizes highlighted a distinct group of German-speaking countries (Germany, Austria, Switzerland), characterized by a preference for organic wholemeal flours \u0026mdash; especially rye and spelt \u0026mdash; whereas Romanian, Finnish, and Italian participants more often preferred non-organic, wheat-based sourdoughs. Northwestern European bakers (Belgium, Denmark, Sweden, UK) commonly used mixed flours and grain blends. Water type and commercial yeast practices also varied by country (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ea).\u003c/p\u003e \u003cp\u003eRegional differences also emerged in baking motivations and perceived health benefits (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eb, Table S9), sourdough storage practices, and household environments (Fig. S6, Table S10). Swiss participants emphasised sustainability, Romanians cited health benefits (digestibility, gut health, glucose control), and Italians rarely mentioned health as a motivation. Enjoyment and taste dominate across countries but are less reported in Belgium. Hierarchical clustering shows shared motivational patterns and sourdough baking experience levels between Swiss and Finnish bakers, while Germany and Italy cluster by a higher proportion of experienced bakers and fewer novices (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eb). In addition to sourdough, other food and beverage fermentations are commonly practiced in France, Austria, and Germany (especially dairy), but rare in Finland (Fig. S6).\u003c/p\u003e \u003cp\u003eDespite patterns related to country of origin and flour preferences, PCA of fermentation features did not yield clustering by country or flour type (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ec,d). Instead, PCA consistently separated household and bakery sourdoughs, pointing in opposite directions in both geographic and flour-related feature spaces (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ec,d). This household-bakery split was mainly driven by higher backslopping frequency, storage temperature, dough size, age, and fermented dairy aromas in bakery sourdoughs (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ec,d). As expected, feeding temperature was associated with shorter fermentation time and dough yield (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ec). Together, these results underline both country- and flour-related trends in how sourdoughs are fermented, while emphasizing individual variation and a strong divide between household and bakery sourdoughs.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThe diversity of processing practices and parameters used in the production of traditional food fermentations plays a significant, yet often understudied, role in their microbiological and chemical properties, environmental sustainability, cultural heritage, and human health effects. Studying the diversity of home fermentation practices at the continental scale presents a non-trivial challenge for conventional scientific research that typically lacks the means to gather such comprehensive data. Citizen Science (CS) offers a solution to this challenge by enabling large-scale sample collection while simultaneously fostering public engagement and enthusiasm for science, strengthening the connection between scientists and the public, and raising awareness about the importance of scientific research related to food fermentation.\u003c/p\u003e \u003cp\u003eThe present study demonstrates the potential of a co-design-driven CS approach to investigate the diversity of technological and cultural parameters of sourdough fermentation across Europe. Through the participation of over 1000 citizen scientists and the collection of more than 670 sourdough samples, a large and geographically diverse dataset was established that would have been difficult to obtain solely through conventional research methods. The co-design-based approach facilitated broad public participation and mutual learning between citizens and researchers, resulting in unique insights into sourdough maintenance motivations and practices, ingredient choices, and perceived health benefits. This approach enabled a comprehensive assessment of sourdough fermentation practices across Europe and permitted validation or reconsideration of prevailing assumptions surrounding traditional sourdough fermentation practices. Furthermore, the project established a sourdough biobank that will serve as a valuable resource for future fermented food research. A key distinction between this study and previous initiatives \u0026mdash; including CS studies like the Global Sourdough Project and experimental studies like Reese \u003cem\u003eet al\u003c/em\u003e. (Landis et al., \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Reese et al., \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) \u0026mdash; is the participatory nature of the HealthFerm citizen science model. Here, participants actively contributed to experiment selection and protocol refinement through feedback loops, making the process more adaptive and participant-centered. While earlier initiatives compiled metadata, the dataset presented in this study includes novel and uniquely detailed information - not only on sourdough fermentation itself, but also on behavioral and cultural dimensions, incorporating comprehensive sensory evaluations and perceived health benefits. Moreover, this co-designed citizen science framework is scalable and transferable to other traditional food systems. It can be tailored to address specific challenges in sample collection or metadata design, underscoring the value of iterative participant engagement in shaping robust research protocols.\u003c/p\u003e \u003cp\u003ePre-launch workshops enhanced engagement by helping to refine sampling kits, instructions, and questionnaires, improving clarity and accessibility. Direct feedback and interactions between citizens and researchers revealed strong participant motivation to contribute to the study, carry out home experiments, and stay informed about project progress and outcomes. These exchanges suggested that hands-on involvement fostered a sense of being valued, sparked curiosity, and strengthened commitment. The collaborative, iterative process appeared to build trust and a sense of ownership, likely sustaining participant interest \u0026mdash; especially in the study outcomes. Future citizen science projects may benefit from emphasizing co-design and transparent researcher\u0026ndash;participant communication to maintain engagement and foster long-term scientific investment.\u003c/p\u003e \u003cp\u003eA significant finding of this study is understanding the public perceptions of the health benefits of sourdough bread. Most participants considered sourdough to be healthier than yeast-leavened bread, commonly citing improved digestibility and gut health as motivating factors for maintaining and baking with sourdough. Public perceptions of health benefits align well with recent scientific findings demonstrating increased mineral bioavailability in sourdough-based products \u003cem\u003ein vitro\u003c/em\u003e, and evidence for improved digestibility or lowering the glycaemic index \u003cem\u003ein vivo\u003c/em\u003e (Arora et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; D\u0026rsquo;Amico et al., \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Ribet et al., \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Rizzello et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Nevertheless, these perceptions reflect the intersection of traditional food fermentation, consumer beliefs, and scientific inquiry, emphasizing the relevance of sourdough in public discussions on health and nutrition.\u003c/p\u003e \u003cp\u003eThe study also faced several challenges. Recruitment biases \u0026mdash; particularly self-selection bias, where volunteers were likely more motivated or interested in fermentation than the general public \u0026mdash; may have skewed participation. Such imbalances are inherent to citizen science, despite broad outreach, and were compounded by logistical challenges, variability in citizen-conducted experiments, and unpredictable sample arrivals, complicating scheduling of microbiological analyses (Berghen et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Constant \u0026amp; Hughes, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Waugh et al., \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eGeographical coverage also presented challenges. Advertisement strategies across a multilingual continent needed persistence and a multifaceted approach, working best when conducted in local languages. Project recruitment and materials were made available in seven languages, reflecting the national languages of the hub labs involved in this project as well as English. Nevertheless, regional language barriers still posed significant obstacles throughout the recruitment, data collection, and communication phases. This most likely explains the disproportionate geographic representation in the project. Countries sharing a language with the research hubs exhibited a high degree of participation (e.g., Switzerland, Germany, and Austria all share German as one of their official languages) whereas countries without language representation in the project yielded low or no participation (e.g., Spanish, Greek, Polish, \u003cem\u003eetc.\u003c/em\u003e). Future citizen science initiatives may benefit from a strong, multilingual social media presence \u0026mdash; potentially supported by digital marketing professionals and enhanced with interactive tools such as chatbots \u0026mdash; to increase visibility, foster engagement, and encourage continued participant involvement.\u003c/p\u003e \u003cp\u003eLogistical barriers also impacted the study. Sampling kit and sourdough sample shipping was most efficient within the European Union (EU), but international shipments crossing non-EU borders encountered issues related to customs regulations and inconsistencies between postal services (availability of prepaid return labels and other services). Additionally, the absence of temperature-controlled shipping resulted in continued fermentation during transit, and prolonged transit times often led to gas buildup, causing some sourdoughs to leak from their containers. To mitigate this issue in future studies, it may be advisable to instruct participants explicitly to ship samples early in the week, thereby reducing the risk of extended exposure to non-cooled conditions during prolonged transit times (e.g., days when postal delivery services are not available). This continued fermentation possibly explains discrepancies between pH measurements taken at home and those recorded upon arrival at the laboratory. Alternatively, this discrepancy could have arisen from the use of colorimetric pH strips with low dynamic range (blue and green for pH 3 to 6) and less precision than the calibrated electronic pH meters used in the hub labs. Follow-up studies are necessary to evaluate potential microbiological and chemical changes that could have arisen from prolonged fermentation without temperature control. Addressing this uncertainty will help refine protocols for sample collection and shipping, ensuring higher data quality in future CS projects where samples are shipped through the postal system.\u003c/p\u003e \u003cp\u003eDespite participant-obtained data often being of similar quality than scientist-obtained data (Lewandowski et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), the large variability in participant-reported data posed significant challenges for curation and interpretation, particularly regarding home experiment results. This variability aligns with findings from citizen science projects, where differences in volunteers\u0026rsquo; skills, motivations, and task familiarity can lead to inconsistencies in data collection (Weigelhofer \u0026amp; P\u0026ouml;lz, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Wiggins \u0026amp; Wilbanks, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Sensory evaluations were conducted using the CATA methodology, a reproducible and consumer-appropriate approach to characterize complex aromas. Here, CATA enabled the collection of large volumes and diverse data. However, the observed variability was high - stemming from both sourdough heterogeneity and individual perceptual differences - which underscores the need for cautious interpretation of citizen-contributed sensory data. Extending workshop formats beyond co-design to include practical sessions prior to performing home experiments could enhance data accuracy and reliability in future CS studies (Fleming et al., \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Piochi et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2021\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eBeyond methodological challenges and lessons learned for future CS initiatives, the study revealed valuable insights into sourdough maintenance practices across Europe. Distinct differences emerged between household and bakery sourdoughs, with bakeries typically maintaining older sourdoughs, backslopping more frequently, and fermenting at higher temperatures to meet the fast-paced demands of professional baking. In addition, the choice of cereal also emerged as a key factor shaping sourdough properties. Rye sourdoughs typically exhibited higher dough yields and total titratable acidity compared to wheat-based sourdoughs. These differences can likely be attributed to the higher arabinoxylan content and enzymatic activity of rye flour, which influences water absorption and fermentation kinetics (Courtin \u0026amp; Delcour, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2002\u003c/span\u003e; Deleu et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Interestingly, the widespread use of organic flour among many participants reflects a strong consumer preference, suggesting that ingredient choices may influence microbial composition in ways that warrant further investigation in follow-up studies.\u003c/p\u003e \u003cp\u003eOne of the most intriguing findings of the study was the emergence of strong regional signatures, where countries exhibited distinct ingredient preferences, fermentation habits, and environmental conditions for maintaining sourdough, as well as differences in baking motivations and perceived health benefits. For instance, German-speaking countries (Germany, Austria, Switzerland) showed a clear inclination toward organic wholemeal flours, particularly rye and spelt, reflecting a robust tradition of non-wheat-based bread making. In contrast, participants from southern and northern Europe primarily used wheat-based sourdoughs and non-organic flours, despite their rich sourdough bread cultures. Interestingly, Finnish participants preferred wheat sourdoughs, despite rye bread\u0026rsquo;s national prominence. This may suggest that rye breads are more commonly purchased than homemade, or it might reflect a latent recruitment bias, where traditional bakers might have been underrepresented or less likely to volunteer for the study. Instead, participants were more likely to be motivated by enjoyment, taste and perceived health benefits rather than tradition, as supported by the fact that fewer than 10% reported baking for traditional reasons, while over 80% cited taste and 70% the enjoyment of baking (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea).\u003c/p\u003e \u003cp\u003eThis study showcases the potential of co-designed citizen science to enable large-scale, multicultural research and generate comprehensive, population-level datasets that can be integrated with microbiome analyses. It also highlights the need for improved strategies to mitigate demographic biases, ensure data reliability, and address logistical challenges, particularly in multilingual and multinational contexts. The widespread belief in sourdough-related health benefits underscores the importance of continued scientific validation, especially regarding gut microbiota and metabolic health (Wiggins \u0026amp; Wilbanks, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Community-generated insights from this study lay a foundation for future investigations into links between sourdough consumption and health outcomes. Moreover, the framework established here offers a scalable model for participatory research, applicable not only to sourdough but also to other traditionally fermented foods. By leveraging citizen participation, future efforts can bridge scientific inquiry with traditional knowledge, advancing our understanding of fermented food microbiomes and their implications for human and planetary health.\u003c/p\u003e"},{"header":"Materials and methods","content":"\u003cp\u003eImplementation and coordination of a multi-hub citizen science study\u003c/p\u003e \u003cp\u003eA co-design-driven CS project embedded in the European HealthFerm project lies at the heart of the current study. From the design phase onward, several key considerations were identified and addressed. These include compliance with key privacy and ethical regulations (such as the General Data Protection Regulation (GDPR), the Nagoya Protocol on Access and Benefit Sharing, and ethical approval), the need for communication in the native languages of the citizen scientists throughout the recruitment, experimental, and dissemination phases, logistical challenges such as sample shipping (\u003cem\u003ee.g.\u003c/em\u003e, shipping time, administration, and costs), and unpredictable sample collection timelines influencing the following laboratory processing workload.\u003c/p\u003e \u003cp\u003eTo manage these complexities, the CS effort was subdivided across five sampling hubs across Europe, each covering a specific geographical area (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The different hubs were each responsible for the GDPR and Nagoya protocol compliance within their collection areas. This subdivision also simplified translation efforts, communication with the participants in seven languages, and the shipping arrangements with local couriers. However, it also introduced challenges in ensuring result comparability, requiring the standardisation of protocols and documentation across all hubs. The sampling hubs were located at the Free University of Bozen-Bolzano (UBZ, Italy), Vrije Universiteit Brussel (VUB, Belgium), Institute of Biology Bucharest (IBB, Romania), University of Helsinki (UH, Finland), and ETH Zurich (ETHZ, Switzerland).\u003c/p\u003e \u003cp\u003eCo-design workshops and questionnaire design\u003c/p\u003e \u003cp\u003eWe chose to follow a co-design process to involve representative participants in the design of research objectives and procedures (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, Fig \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003eb). These were further refined through multidisciplinary collaboration within the HealthFerm consortium, drawing on expertise in sourdough research, citizen science, microbiome science, food technology, food biochemistry, consumer science, and sensory analysis. To support the co-design process, we prepared draft materials essential for the campaign, including a sourdough metadata questionnaire, an instruction booklet for sampling and at-home experiments, and a sampling kit. Feedback on these prototypes was collected through three co-design workshops held in Belgium, Finland, and Switzerland, involving small groups of citizen scientists representative of the target audience - namely home and small-scale commercial sourdough bakers. All workshops followed a similar structure, beginning with an introductory presentation on the HealthFerm project and the CS initiative, followed by an interactive, layman-friendly presentation on sourdough and the current state of sourdough research, designed to both assess participants\u0026rsquo; knowledge levels and to provide them with a foundational understanding of the science behind sourdough. This was followed by an interactive session where participants were guided through the draft questionnaire and instructions, while also trying some proposals for at-home experiments, including a pH measurement and a sensory analysis of a provided sourdough starter. Feedback from the workshop participants was collected throughout the session. The responses obtained were compared across the different workshops and key insights were incorporated in the further development of the sampling kits and sourdough metadata questionnaire used during the subsequent sampling campaign.\u003c/p\u003e \u003cp\u003eParticipant registration and at-home experimentation\u003c/p\u003e \u003cp\u003eThe citizen science (CS) sourdough campaign was launched via an online sign-up form (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003eb), accessed through a multilingual registration page on the HealthFerm website (available in English and the six official languages of the participating hub countries). Participants were automatically assigned to a local hub laboratory based on location. Each hub hosted its own language-specific, GDPR-compliant registration survey - including informed consent, pseudonymization, and a sample rights transfer agreement - using Microsoft Forms or Google Forms, depending on institutional preferences.\u003c/p\u003e \u003cp\u003eThe survey included 90 questions, collecting only essential personal information (e.g., name, email, address) for kit delivery and metadata on the participant\u0026rsquo;s sourdough. Respondents could also indicate willingness to contribute sourdough or other fermented food samples.\u003c/p\u003e \u003cp\u003eRecruitment was promoted via social media, press coverage, print and digital ads, university networks, public lectures, and conferences. Outreach also targeted sourdough communities, fermentation networks, professional bakers, and influencers. A dedicated HealthFerm CS Blog (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://healthferm.eu/news-and-events/blog\u003c/span\u003e\u003cspan address=\"https://healthferm.eu/news-and-events/blog\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e)\u003c/span\u003e and FAQ (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://healthferm.eu/healthferm-\u003c/span\u003e\u003cspan address=\"https://healthferm.eu/healthferm-\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003cp\u003e \u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003ecommunity/faq\u003c/span\u003e) were maintained on the project website, with regular updates via LinkedIn and newsletters.\u003c/p\u003e \u003cp\u003eTo ensure diverse geographic and substrate representation, participants were sub-selected for sample submission. Selected participants received a sampling kit by mail containing instructions, materials for sourdough sampling and home experiments (Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e), and a unique five-character code for pseudonymization. A QR code linked to a follow-up survey (using the same secure tools) to upload experiment results and updated metadata.\u003c/p\u003e \u003cp\u003eParticipants followed a standardized protocol (archived on Zenodo; see Data availability) to analyze their own sourdoughs and breads. Sourdough pH was measured by mixing 10 g of dough with 10 g of water and using pH strips (MilliPoreSigma, Darmstadt, Germany). Bread pH was assessed by homogenizing 10 g of crumb (crust removed) in 100 g of water. Bread density was determined by measuring a\u0026thinsp;~\u0026thinsp;3 cm crustless slice\u0026rsquo;s weight and dimensions. Aroma descriptions of sourdough and bread were recorded using sensory wheels developed by the University of Copenhagen for the Wild Sourdough and Smag for Livet projects (\u003cem\u003eLav Din Egen Surdej \u0026ndash; M\u0026aring;l Og Vurd\u0026eacute;r | Smagforlivet.Dk\u003c/em\u003e, n.d.; Nichols, n.d.)\u003c/p\u003e \u003cp\u003eSourdough sample collection and processing\u003c/p\u003e \u003cp\u003eSourdough samples were sent in ambient conditions by mail for lab analyses. Upon reception, samples were processed within 24 hours at the local collection hubs using common protocols and chemicals. The pH was determined by diluting 10 g of sourdough in 100 mL of deionised water. The same suspension was used to assess the total titratable acidity (TTA) by titrating with 0.1 M NaOH until a pH of 8.5 was reached, with the volume of NaOH added representing the TTA value. For future analyses and biobanking, 5 g of sample was diluted in either 5 g of sterile deionised water or 5 g of sterile 50% glycerol, aliquoted into cryovials, and stored at -80\u0026deg;C.\u003c/p\u003e \u003cp\u003eSurvey processing, statistics, and data visualization\u003c/p\u003e \u003cp\u003eEach hub maintained three multilingual datasets: registration surveys, result surveys, and laboratory measurements. After sample collection, pseudo-anonymised datasets were centrally compiled, translated into English (semi-automated, using Google Sheets), and merged into unified registration, result, and laboratory tables. Manual curation was used to standardise vocabulary and harmonise free-text entries. Datasets were linked via unique 5-character sample codes and person IDs (archived on Zenodo; see Data availability); missing result variables were filled from registration data when available.\u003c/p\u003e \u003cp\u003eSome metadata - such as sensory attributes, pH, and bread density - were only collected during the sample phase and are thus available for a subset of samples. Flour types were classified from participant-reported grain and milling grade. In this study, \u003cem\u003eTriticum aestivum\u003c/em\u003e is referred to as \"wheat,\" \u003cem\u003eTriticum spelta\u003c/em\u003e as \"spelt,\" and \u003cem\u003eSecale cereale\u003c/em\u003e as \"rye\". For chi-square analyses, semi-wholemeal and wholemeal flours were grouped (as semi-wholemeal was only available for wheat and clustered closely with wholemeal wheat in both lab and feeding-parameter data). For selected enrichment analyses, composite flour types (e.g., wholemeal wheat) were decomposed into grain base and milling grade to distinguish associations with participant region, motivation, or health perception. In contrast, numerical and feature-space analyses retained these as separate categories to preserve resolution. Descriptive and correlative analyses excluded missing data on a per-variable basis unless stated otherwise. Data integration, curation, and analysis were conducted using Python 3.12.7 and R 4.4.1 (R Core Team, 2021).\u003c/p\u003e \u003cp\u003e \u003cb\u003eDescriptive analyses\u003c/b\u003e. Descriptive analysis based on a bakery-household distinction was performed using the dplyr, tidyr, and ggplot2 packages in R. Data were filtered to exclude the samples with an unknown sample source, along with one single sample of industrial origin. Categorisation was introduced for several of the used metadata (dough age, fermentation temperature, fermentation time, storage temperature, and dough yield) by which values present at \u0026le;\u0026thinsp;1% prevalence in each category were binned as \u0026lsquo;Other\u0026rsquo;.\u003c/p\u003e \u003cp\u003e \u003cb\u003ePairwise analysis\u003c/b\u003e. Both t-tests and the Mann-Whitney U tests were used depending on the data\u0026rsquo;s normality and homoscedasticity, while chi-square tests were applied for binary yes/no responses and categorical variables. Effect sizes were calculated using the formula for Cohen\u0026rsquo;s d, rank serial correlation, and Cram\u0026eacute;r\u0026rsquo;s V for t-tests, Mann-Whitney U test, and chi-square tests, respectively.\u003c/p\u003e \u003cp\u003e \u003cb\u003eMaps.\u003c/b\u003e To visualise the geographic distribution of sourdough samples, spatial data processing and mapping were conducted using Python (v3.10.14), incorporating geopandas (v1.0.1) for geospatial data handling, shapely (v2.0.6) for geometric operations, matplotlib (v3.8.4) for data visualization, and contextility (v1.6.2) for basemap integration.\u003c/p\u003e \u003cp\u003e\u003cb\u003eCollinearity analysis.\u003c/b\u003e Collinearity among variables was assessed prior to multivariate and enrichment analysis. Categorical associations were tested using Chi-square and Cram\u0026eacute;r\u0026rsquo;s V; numerical variables using Pearson, Spearman, and cosine similarity; and mixed types with Kruskal\u0026ndash;Wallis tests. False discovery rate was controlled using the Benjamini\u0026ndash;Hochberg method. Hierarchical clustering of association matrices guided the selection of representative variables to avoid redundancy and inclusion of collinear variables.\u003c/p\u003e \u003cp\u003e \u003cb\u003eEnrichment analyses by category\u003c/b\u003e. Due to higher sampling density and geographic representation in Europe, geolocation-based enrichment analyses excluded non-European samples (longitude \u0026minus;\u0026thinsp;50 to 70, latitude 0 to 70). To assess categorical enrichment across geographic clusters (e.g., countries), Chi-square tests were performed for each category\u0026ndash;cluster pair. For significant associations, Cram\u0026eacute;r\u0026rsquo;s V quantified effect size, and enrichment scores indicated over- or underrepresentation. P-values were adjusted using Benjamini\u0026ndash;Hochberg false discovery rate (BH-FDR). For numerical features, Mann\u0026ndash;Whitney U tests were applied to each feature\u0026ndash;cluster pair, comparing within-cluster values to all others. Enrichment scores were derived from mean differences, with a small constant (ϵ = 1\u0026times;10⁻⁴) added to avoid division by zero. Scores were log₂-normalised for scale-independent comparisons. BH-FDR-adjusted p-values were used to classify features as uniquely or multiply enriched. Visualisation followed the same clustering strategy, with only significant log₂-normalised enrichments displayed. For both categorical and numerical enrichments, positive enrichments (PE) and negative enrichments (NE) and uniquely (U) or multiply (M) associated features with clusters were flagged in output tables (Table S2; S3; S8-10) and visualised accordingly.\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:Category\\:Enrichment\\:Score\\:=\\:\\frac{Count\\:of\\:category\\:in\\:cluster/Total\\:count\\:in\\:cluster}{Total\\:count\\:of\\:category/Total\\:samples}$$\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Equb\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equb\" name=\"EquationSource\"\u003e\n$$\\:Numerical\\:Enrichment\\:Score\\:=\\:\\frac{Variable\\:mean\\:in\\:cluster\\:+\\:\\epsilon\\:}{Variable\\:mean\\:overall\\:+\\:\\epsilon\\:}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003e \u003cb\u003eAnalysis of sourdough aromas.\u003c/b\u003e Sensory diversity was assessed using Jaccard distances (full aroma spectrum). PERMANOVA was applied to the distance matrix to test categorical metadata effects, including only variables with \u0026ge;\u0026thinsp;3 samples and applying BH-FDR correction. Relationships between sensory dissimilarities and numerical metadata were evaluated using Mantel tests (Pearson and Spearman) on scaled variables, with dynamic exclusion of missing data. Principal Coordinate Analysis (PCoA) was performed for visualisation, retaining the first two axes (PC1, PC2). The top 10 aroma features explaining variance were identified via Pearson correlations with principal coordinates. Significant metadata were integrated into the PCoA plot, with scaled feature loadings shown as arrows.\u003c/p\u003e \u003cp\u003e \u003cb\u003eAnalysis of numerical metadata feature space.\u003c/b\u003e Principal Component Analysis (PCA) was performed using scikit-learn. Only non-collinear variables were considered; samples with missing values for the selected variables were excluded. PCA biplots were generated in matplotlib with confidence ellipses added at 95%.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAcknowledgements\u003c/h2\u003e\n\u003cp\u003eThe authors acknowledge financial support from the project HealthFerm, which is funded by the European Union under the Horizon Europe grant agreement No. 101060247 and by the Swiss State Secretariat for Education, Research and Innovation (SERI) under contract No. 22.00210. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union nor European Research Executive Agency (REA). Neither the European Union nor REA can be held responsible for them.\u003c/p\u003e\n\u003cp\u003eThe authors thank Alessandro Rearte, Rosy Mondardini, Wannes De Man, Michael Bom Fr\u0026oslash;st, Vimac Nolla Ard\u0026egrave;vol, Jean-Paul Garin, Elena Martucci, and the rest of the HealthFerm consortium for their advice and support in setting up, promoting and advertising the project. We also thank Luisa Ferreira, Doriela Grabocka, Celine Verdonck, Silvia-Simona Grosu-Tudor, and Emanuela-Catalina Ionetic for their help in data curation and/or technical support. We offer our gratitude to all the participants who attended the co-design workshops, completed the questionnaires, submitted their sourdoughs, performed home experiments, and patiently waited for their results. None of this would have been possible without your contributions.\u003c/p\u003e\n\u003ch2\u003eAuthor information\u003c/h2\u003e\n\u003ch2\u003eContributions\u003c/h2\u003e\n\u003cp\u003eConceptualization: N.A.B., S.W., C.M.C., and R.C. conceived and designed the study with input from other authors. Methodology: A.M., T.G., J.P.T., F.C., Y.D.B., and N.A.B. designed the survey and experimental methodology with input from other authors. Investigation: A.M., T.G., J.P.T., C.V., I.-R.A., M.N., and A.Z.A.T. carried out data collection. Curation: A.M. and J.P.T. curated and managed the data. Formal analysis: A.M., J.P.T., and T.G. Visualization: A.M., T.G., and J.P.T. Supervision: N.A.B., L.D., M.G., M.Z., R.C., L.N., S.W., and C.M.C. Writing \u0026mdash; original draft: A.M., T.G., J.P.T., and N.A.B. All authors contributed to the review and editing of the manuscript and approved the final version.\u003c/p\u003e\n\u003ch2\u003eEthics declarations\u003c/h2\u003e\n\u003cp\u003eThis study obtained ethical approval for the activities of each respective hub lab. Approval was granted by the ethics committees of the respective universities under the reference numbers HealthFerm_Cod2023_14 (UniBZ) , ECHW_429 (VUB), 183/2022 (IBB), 52/2023 (UH), and EK 2023-N-118 (ETHZ).\u003c/p\u003e\n\u003ch2\u003eCompeting interests\u003c/h2\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\n\u003ch2\u003eData availability\u003c/h2\u003e\n\u003cp\u003eThe dataset supporting this study \u0026mdash; including questionnaire files (Survey 1 and Survey 2), the home experiment instruction booklet, and the fully integrated metadata \u0026mdash; is currently under restricted access to preserve the integrity of peer review. It will be made publicly available under Creative Commons Attribution 4.0 International (CC BY 4.0) license upon article publication. In the meantime, it is available upon request via Zenodo:\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;https://doi.org/10.5281/zenodo.15366921.\u003c/p\u003e\n\u003ch2\u003eCode availability\u003c/h2\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eArora, K., Ameur, H., Polo, A., Di Cagno, R., Rizzello, C. G., \u0026amp; Gobbetti, M. (2021). Thirty years of knowledge on sourdough fermentation: A systematic review. \u003cem\u003eTrends in Food Science \u0026amp; Technology\u003c/em\u003e, \u003cem\u003e108\u003c/em\u003e, 71\u0026ndash;83. https://doi.org/10.1016/j.tifs.2020.12.008\u003c/li\u003e\n\u003cli\u003eBerghen, B. V., Vanermen, I., \u0026amp; Vranken, L. (2024). Citizen scientists: Unveiling motivations and characteristics influencing initial and sustained participation in an agricultural project. \u003cem\u003ePLOS ONE\u003c/em\u003e, \u003cem\u003e19\u003c/em\u003e(5), e0303103. https://doi.org/10.1371/journal.pone.0303103\u003c/li\u003e\n\u003cli\u003eBondt, Y. D., Verdonck, C., Brandt, M. J., Vuyst, L. D., G\u0026auml;nzle, M. G., Gobbetti, M., Zannini, E., \u0026amp; Courtin, C. M. (2024). Wheat Sourdough Breadmaking: A Scoping Review. \u003cem\u003eAnnual Review of Food Science and Technology\u003c/em\u003e, \u003cem\u003e15\u003c/em\u003e(Volume 15, 2024), 265\u0026ndash;282. https://doi.org/10.1146/annurev-food-110923-034834\u003c/li\u003e\n\u003cli\u003eCalvert, M. D., Madden, A. A., Nichols, L. M., Haddad, N. M., Lahne, J., Dunn, R. R., \u0026amp; McKenney, E. A. (2021). A review of sourdough starters: Ecology, practices, and sensory quality with applications for baking and recommendations for future research. \u003cem\u003ePeerJ\u003c/em\u003e, \u003cem\u003e9\u003c/em\u003e, e11389. https://doi.org/10.7717/peerj.11389\u003c/li\u003e\n\u003cli\u003eCappelle, S., Guylaine, L., \u0026amp; Gobbetti, M. (2013). History and Social Aspects of Sourdough. In \u003cem\u003eHandbook on Sourdough Biotechnology\u003c/em\u003e (pp. 1\u0026ndash;10). https://doi.org/10.1007/978-1-4614-5425-0_1\u003c/li\u003e\n\u003cli\u003eCavicchioli, R., Ripple, W. J., Timmis, K. N., Azam, F., Bakken, L. R., Baylis, M., Behrenfeld, M. J., Boetius, A., Boyd, P. W., Classen, A. T., Crowther, T. W., Danovaro, R., Foreman, C. M., Huisman, J., Hutchins, D. A., Jansson, J. K., Karl, D. M., Koskella, B., Mark Welch, D. B., \u0026hellip; Webster, N. S. (2019). Scientists\u0026rsquo; warning to humanity: Microorganisms and climate change. \u003cem\u003eNature Reviews Microbiology\u003c/em\u003e, \u003cem\u003e17\u003c/em\u003e(9), 569\u0026ndash;586. https://doi.org/10.1038/s41579-019-0222-5\u003c/li\u003e\n\u003cli\u003eComasio, A., Verce, M., Van Kerrebroeck, S., \u0026amp; De Vuyst, L. (2020). Diverse Microbial Composition of Sourdoughs From Different Origins. \u003cem\u003eFrontiers in Microbiology\u003c/em\u003e, \u003cem\u003e11\u003c/em\u003e. https://www.frontiersin.org/journals/microbiology/articles/10.3389/fmicb.2020.01212\u003c/li\u003e\n\u003cli\u003eConstant, N. L., \u0026amp; Hughes, J. (2023). Diversifying citizen science through the inclusion of young people. \u003cem\u003eJournal of Science Communication\u003c/em\u003e, \u003cem\u003e22\u003c/em\u003e(2), A07. https://doi.org/10.22323/2.22020207\u003c/li\u003e\n\u003cli\u003eCourtin, C. M., \u0026amp; Delcour, J. A. (2002). Arabinoxylans and Endoxylanases in Wheat Flour Bread-making. \u003cem\u003eJournal of Cereal Science\u003c/em\u003e, \u003cem\u003e35\u003c/em\u003e(3), 225\u0026ndash;243. https://doi.org/10.1006/jcrs.2001.0433\u003c/li\u003e\n\u003cli\u003eD\u0026rsquo;Amico, V., G\u0026auml;nzle, M., Call, L., Zwirzitz, B., Grausgruber, H., D\u0026rsquo;Amico, S., \u0026amp; Brouns, F. (2023). Does sourdough bread provide clinically relevant health benefits? \u003cem\u003eFrontiers in Nutrition\u003c/em\u003e, \u003cem\u003eVolume 10-2023\u003c/em\u003e. https://www.frontiersin.org/journals/nutrition/articles/10.3389/fnut.2023.1230043\u003c/li\u003e\n\u003cli\u003eDe Vuyst, L., Comasio ,Andrea, \u0026amp; and Kerrebroeck, S. V. (2023). Sourdough production: Fermentation strategies, microbial ecology, and use of non-flour ingredients. \u003cem\u003eCritical Reviews in Food Science and Nutrition\u003c/em\u003e, \u003cem\u003e63\u003c/em\u003e(15), 2447\u0026ndash;2479. https://doi.org/10.1080/10408398.2021.1976100\u003c/li\u003e\n\u003cli\u003eDe Vuyst, L., Van Kerrebroeck, S., \u0026amp; Leroy, F. (2017). Chapter Two\u0026mdash;Microbial Ecology and Process Technology of Sourdough Fermentation. In S. Sariaslani \u0026amp; G. M. Gadd (Eds.), \u003cem\u003eAdvances in Applied Microbiology\u003c/em\u003e (Vol. 100, pp. 49\u0026ndash;160). Academic Press. https://doi.org/10.1016/bs.aambs.2017.02.003\u003c/li\u003e\n\u003cli\u003eDeleu, L. J., Lemmens, E., Redant, L., \u0026amp; Delcour, J. A. (2020). The major constituents of rye (Secale cereale L.) flour and their role in the production of rye bread, a food product to which a multitude of health aspects are ascribed. \u003cem\u003eCereal Chemistry\u003c/em\u003e, \u003cem\u003e97\u003c/em\u003e(4), 739\u0026ndash;754. https://doi.org/10.1002/cche.10306\u003c/li\u003e\n\u003cli\u003eFleming, E. E., Ziegler, G. R., \u0026amp; Hayes, J. E. (2015). Check-all-that-apply (CATA), sorting, and polarized sensory positioning (PSP) with astringent stimuli. \u003cem\u003eFood Quality and Preference\u003c/em\u003e, \u003cem\u003e45\u003c/em\u003e, 41\u0026ndash;49. https://doi.org/10.1016/j.foodqual.2015.05.004\u003c/li\u003e\n\u003cli\u003eFraisl, D., Hager, G., Bedessem, B., Gold, M., Hsing, P.-Y., Danielsen, F., Hitchcock, C. B., Hulbert, J. M., Piera, J., Spiers, H., Thiel, M., \u0026amp; Haklay, M. (2022). Citizen science in environmental and ecological sciences. \u003cem\u003eNature Reviews Methods Primers\u003c/em\u003e, \u003cem\u003e2\u003c/em\u003e(1), 1\u0026ndash;20. https://doi.org/10.1038/s43586-022-00144-4\u003c/li\u003e\n\u003cli\u003eGarcia, S., Ordo\u0026ntilde;ez, S., L\u0026oacute;pez-Molina, V. M., Lacruz-Pleguezuelos, B., Carrillo de Santa Pau, E., \u0026amp; Marcos-Zambrano, L. J. (2023). Citizen science helps to raise awareness about gut microbiome health in people at risk of developing non-communicable diseases. \u003cem\u003eGut Microbes\u003c/em\u003e, \u003cem\u003e15\u003c/em\u003e(1), 2241207. https://doi.org/10.1080/19490976.2023.2241207\u003c/li\u003e\n\u003cli\u003eGupta, A., Gupta, R., \u0026amp; Singh, R. L. (2016). Microbes and Environment. \u003cem\u003ePrinciples and Applications of Environmental Biotechnology for a Sustainable Future\u003c/em\u003e, 43. https://doi.org/10.1007/978-981-10-1866-4_3\u003c/li\u003e\n\u003cli\u003eHern\u0026aacute;ndez-Vel\u0026aacute;zquez, R., Fl\u0026ouml;rl, L., Lavrinienko, A., Sebechlebsk\u0026aacute;, Z., Merk, L., Greppi, A., \u0026amp; Bokulich, N. A. (2024). The future is fermented: Microbial biodiversity of fermented foods is a critical resource for food innovation and human health. \u003cem\u003eTrends in Food Science \u0026amp; Technology\u003c/em\u003e, \u003cem\u003e150\u003c/em\u003e, 104569. https://doi.org/10.1016/j.tifs.2024.104569\u003c/li\u003e\n\u003cli\u003eLandis, E. A., Oliverio, A. M., McKenney, E. A., Nichols, L. M., Kfoury, N., Biango-Daniels, M., Shell, L. K., Madden, A. A., Shapiro, L., Sakunala, S., Drake, K., Robbat, A., Booker, M., Dunn, R. R., Fierer, N., \u0026amp; Wolfe, B. E. (2021). The diversity and function of sourdough starter microbiomes. \u003cem\u003eELife\u003c/em\u003e, \u003cem\u003e10\u003c/em\u003e, e61644. https://doi.org/10.7554/eLife.61644\u003c/li\u003e\n\u003cli\u003eLewandowski, E., Caldwell, W., Elmquist, D., \u0026amp; Oberhauser, K. (2017). Public Perceptions of Citizen Science. \u003cem\u003eCitizen Science: Theory and Practice\u003c/em\u003e, \u003cem\u003e2\u003c/em\u003e(1). https://doi.org/10.5334/cstp.77\u003c/li\u003e\n\u003cli\u003eMarco, M. L., Sanders, M. E., G\u0026auml;nzle, M., Arrieta, M. C., Cotter, P. D., De Vuyst, L., Hill, C., Holzapfel, W., Lebeer, S., Merenstein, D., Reid, G., Wolfe, B. E., \u0026amp; Hutkins, R. (2021). The International Scientific Association for Probiotics and Prebiotics (ISAPP) consensus statement on fermented foods. \u003cem\u003eNature Reviews Gastroenterology \u0026amp; Hepatology\u003c/em\u003e, \u003cem\u003e18\u003c/em\u003e(3), 196\u0026ndash;208. https://doi.org/10.1038/s41575-020-00390-5\u003c/li\u003e\n\u003cli\u003ePateman, R. M., de Bruin, A., Piirsalu, E., Reynolds, C., Stokeld, E., \u0026amp; West, S. E. (2020). Citizen Science for Quantifying and Reducing Food Loss and Food Waste. \u003cem\u003eFrontiers in Sustainable Food Systems\u003c/em\u003e, \u003cem\u003e4\u003c/em\u003e. https://doi.org/10.3389/fsufs.2020.589089\u003c/li\u003e\n\u003cli\u003ePiochi, M., Cabrino, G., \u0026amp; Torri, L. (2021). Check-All-That-Apply (CATA) Test to Investigate the Consumers\u0026rsquo; Perception of Olive Oil Sensory Properties: Effect of Storage Time and Packaging Material. \u003cem\u003eFoods\u003c/em\u003e, \u003cem\u003e10\u003c/em\u003e(7), 1551. https://doi.org/10.3390/foods10071551\u003c/li\u003e\n\u003cli\u003eReese, A. T., Madden, A. A., Joossens, M., Lacaze, G., \u0026amp; Dunn, R. R. (2020). Influences of Ingredients and Bakers on the Bacteria and Fungi in Sourdough Starters and Bread. \u003cem\u003eMSphere\u003c/em\u003e, \u003cem\u003e5\u003c/em\u003e(1), 10.1128/msphere.00950-19. https://doi.org/10.1128/msphere.00950-19\u003c/li\u003e\n\u003cli\u003eRibet, L., Dessalles, R., Lesens, C., Brusselaers, N., \u0026amp; Durand-Dubief, M. (2023). Nutritional benefits of sourdoughs: A systematic review. \u003cem\u003eAdvances in Nutrition\u003c/em\u003e, \u003cem\u003e14\u003c/em\u003e(1), 22\u0026ndash;29. https://doi.org/10.1016/j.advnut.2022.10.003\u003c/li\u003e\n\u003cli\u003eRizzello, C. G., Portincasa, P., Montemurro, M., Di Palo, D. M., Lorusso, M. P., De Angelis, M., Bonfrate, L., Genot, B., \u0026amp; Gobbetti, M. (2019). Sourdough Fermented Breads are More Digestible than Those Started with Baker\u0026rsquo;s Yeast Alone: An In Vivo Challenge Dissecting Distinct Gastrointestinal Responses. \u003cem\u003eNutrients\u003c/em\u003e, \u003cem\u003e11\u003c/em\u003e(12), 2954. https://doi.org/10.3390/nu11122954\u003c/li\u003e\n\u003cli\u003eValentino, V., Magliulo, R., Farsi, D., Cotter, P. D., O\u0026rsquo;Sullivan, O., Ercolini, D., \u0026amp; De Filippis, F. (2024). Fermented foods, their microbiome and its potential in boosting human health. \u003cem\u003eMicrobial Biotechnology\u003c/em\u003e, \u003cem\u003e17\u003c/em\u003e(2), e14428. https://doi.org/10.1111/1751-7915.14428\u003c/li\u003e\n\u003cli\u003eWalsh, L. H., Breselge, S., Martin, J. G. P., Coakley, M., Ferguson, E., Stapleton, A., Crispie, F., O\u0026rsquo;Toole, P. W., \u0026amp; Cotter, P. D. (2024). Kefir4All, a citizen science initiative to raise awareness of the roles that microbes play in food fermentation. \u003cem\u003eJournal of Microbiology \u0026amp; Biology Education\u003c/em\u003e, \u003cem\u003e0\u003c/em\u003e(0), e00155-23. https://doi.org/10.1128/jmbe.00155-23\u003c/li\u003e\n\u003cli\u003eWaugh, J. K., Lindsey, J. K., Stewart, M. Z., Winter, J. C., \u0026amp; Parrish, J. K. (2023). Demographics of Public Participation in Science: A Meta-Analytic Approach. \u003cem\u003eCitizen Science: Theory and Practice\u003c/em\u003e, \u003cem\u003e8\u003c/em\u003e(1). https://doi.org/10.5334/cstp.610\u003c/li\u003e\n\u003cli\u003eWeigelhofer, G., \u0026amp; P\u0026ouml;lz, E.-M. (2016). Data Quality in Citizen Science Projects: Challenges and Solutions. \u003cem\u003eFrontiers in Environmental Science\u003c/em\u003e, \u003cem\u003e4\u003c/em\u003e. https://doi.org/10.3389/conf.FENVS.2016.01.00011\u003c/li\u003e\n\u003cli\u003eWiggins, A., \u0026amp; Wilbanks, J. (2019). The Rise of Citizen Science in Health and Biomedical Research. \u003cem\u003eThe American Journal of Bioethics\u003c/em\u003e, \u003cem\u003e19\u003c/em\u003e(8), 3\u0026ndash;14. https://doi.org/10.1080/15265161.2019.1619859\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"npj-science-of-food","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"npjscifood","sideBox":"Learn more about [npj Science of Food](http://www.nature.com/npjscifood/)","snPcode":"41538","submissionUrl":"https://submission.springernature.com/new-submission/41538/3","title":"npj Science of Food","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"NPJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-6627352/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6627352/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eFermented foods are culturally significant and increasingly recognized for their potential health benefits, yet scientific data on household fermentation practices remain limited. We launched a co-designed citizen science (CS) initiative within the HealthFerm project to collect sourdough fermentation data across Europe. Over 1000 participants from 33 countries registered, with 671 samples submitted, enabling large-scale analysis of fermentation practices, motivations, and sourdough characteristics. Participants also completed standardized at-home experiments and sensory evaluations, generating a dataset linking baking habits with physicochemical and sensory profiles. Distinct patterns emerged: professional bakers used older, more frequently refreshed starters and fermented at higher temperatures. Ingredient choices and motivations varied by country, shaped by perceived health benefits. Beyond data collection, this initiative established a microbial biobank and harmonized metadata resource, while offering practical insights into co-design, logistics, and public engagement. The resulting framework provides a transferable model for participatory research in microbiology and food systems science.\u003c/p\u003e","manuscriptTitle":"Rising together: Exploring Sourdough Fermentation Diversity through Co-design in the HealthFerm Citizen Science Initiative","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-06-04 06:00:41","doi":"10.21203/rs.3.rs-6627352/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"npj-science-of-food","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"npjscifood","sideBox":"Learn more about [npj Science of Food](http://www.nature.com/npjscifood/)","snPcode":"41538","submissionUrl":"https://submission.springernature.com/new-submission/41538/3","title":"npj Science of Food","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"NPJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"8057e1d1-f670-4a89-aebc-fbf8289cab76","owner":[],"postedDate":"June 4th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":49298227,"name":"Social science/Science technology and society"},{"id":49298228,"name":"Biological sciences/Microbiology/Applied microbiology"}],"tags":[],"updatedAt":"2026-02-02T16:02:18+00:00","versionOfRecord":{"articleIdentity":"rs-6627352","link":"https://doi.org/10.1038/s41538-026-00725-7","journal":{"identity":"npj-science-of-food","isVorOnly":false,"title":"npj Science of Food"},"publishedOn":"2026-01-27 15:58:41","publishedOnDateReadable":"January 27th, 2026"},"versionCreatedAt":"2025-06-04 06:00:41","video":"","vorDoi":"10.1038/s41538-026-00725-7","vorDoiUrl":"https://doi.org/10.1038/s41538-026-00725-7","workflowStages":[]},"version":"v1","identity":"rs-6627352","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6627352","identity":"rs-6627352","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}