This project is not a formally defined research project in the academic sense. Therefore, it was not carried out in accordance with the highest scientific standards, such as those that would apply to peer-reviewed studies. Nevertheless, great importance was attached to accuracy, transparency and traceability so that interested parties can easily understand the data basis, the classifications and the methods used and, if desired, critically examine them. I am happy to answer any questions or comments.
Data was collected from 3 July 2024 to 2 July 2025. During this period, notes were taken several times a day, mostly immediately after each meal. Once a day, the data was transferred in a structured manner to an Excel spreadsheet.
All clearly identifiable plants that appeared in food or beverages were recorded. This included whole plant organs as well as parts of organs (e.g. coconut endosperm), plant extracts (e.g. tea) and isolated plant ingredients (e.g. sucrose from sugar cane or sugar beet). All foods prepared at home, purchased or consumed in restaurants were recorded. Plant ingredients in medicines or cosmetic products were not included, nor were those listed under statements such as ‘may contain traces of ...’.
The data was recorded in binary form: only whether a plant was consumed on a given day or not was documented. Quantities or intra-daily frequencies were not included in the analysis. No distinction was made between different forms of processing of the same plant; for example, table grapes, raisins and wine were uniformly assigned to the species Vitis vinifera.
During the observation period, 222 clearly distinguishable plants were recorded. These are botanically distinguishable units as they appeared in the food. The 206 associated species result from the aggregation of these units at species level.
Cultivated forms of a species were recorded separately if they were clearly distinguishable morphologically or functionally (e.g. broccoli, red cabbage and Brussels sprouts: Brassica oleracea). Different cultivars of a fruit (e.g. apple varieties) were not counted separately.
The taxonomic classification is based on two established references:
Strasburger – Lehrbuch der Pflanzenwissenschaften (38th edition)
Stevens, P. F.: Angiosperm Phylogeny Website (APWeb), Version 14 (2017) and ongoing updates
http://www.mobot.org/MOBOT/research/APweb/
Synonyms have been harmonised; all classifications are consistent at the order, family, genus and species levels.
Whether two plants are considered species, subspecies or varieties is sometimes controversial and subject to scientific revision. This applies in particular to cultivated crops such as wheat (Triticum aestivum) and spelt (Triticum aestivum ssp. spelta). The classification follows the scientific understanding at the time of recording, but takes into account that later reassessments may necessitate adjustments.
The classification according to usage groups is primarily based on Lieberei & Reissdorf, Nutzpflanzen (Crop Plants). Missing or ambiguous cases were supplemented at our discretion. For plants with several typical uses, the classification was determined by the predominant use in personal consumption.
The data was processed in Excel and Power Query. The following was carried out:
Cleaning up inconsistent spellings, harmonising botanical names, plausibility checks, preparing the data structures for statistical evaluation and web visualisation.
The Shannon index and the Simpson index were used to evaluate plant diversity. Both measures take into account not only the number of different elements, but also their relative distribution.
For each plant that was consumed on at least ten days during the observation period, the months in which it occurred and how its occurrence was distributed across the seasons were determined.
Co-occurrence, support, confidence and lift were calculated to investigate plants that are typically consumed together. These measures allow conclusions to be drawn about which combinations occur more or less frequently than would be expected by chance. Association pairs with a support <0.05 were not taken into account.
The number of angiosperm orders represented in the data set was determined, as well as how the observed plants are distributed across these orders.
This study is based on a single-person data set that reflects individual consumption habits over the course of one year and therefore cannot claim to be universally applicable. Since only the occurrence of a plant per day was recorded and not its quantity, no quantitative statements can be made about consumption intensities.
Subjective decisions were unavoidable, especially when distinguishing between cultivated forms, subspecies and usage groups. Complex dishes with unknown or unidentifiable ingredients may have led to simplified classifications.
In addition, botanical taxonomy can change over time. New findings may lead to taxonomic categories used today being defined differently in the future. All classifications are therefore snapshots of the state of knowledge at the time of recording.
