Microcosms based on different breeding substrates of Drosophila melanogaster (apple, banana, grape, plum, lemon, onion, tomato) were set up from pieces of substrate colonized by insect associated microbes in the field (Riedel & Rohlfs, in preparation). Microbes were maintained by Drosophila melanogaster through ingestion and reinoculation by fecal droplet deposition on fresh substrates provided ad libitum. DNA was extracted from fecal material of groups of female flies and from substrates after fly development in an experimental set-up. Fungal and bacterial amplicon regions were used to identify fly associated microorganism from said fecal material using the primers 341F and 785R (Klindworth et al., 2013) spanning the V3-V4 region of the 16S ribosomal RNA gene, and the primers ITS3_KYO2 (forward) and ITS4_KY03 (reverse) covering the Internal spacer region two (ITS2, Toju et al., 2012). The raw demultiplexed sequences were created on an Illumina MiSeq v3.0 (2 × 300-bp) run conducted at Advanced Identification Methods (AIM), Leipzig, Germany. Related material samples of fly life-history trait data, i.e. development time and weight of flies are included. These data stem from axenic larvae transferred into individual developmental environments of either allochthonous or autochthonous substrate-microbiota conditions, including fecal material from flies kept in apple, banana and tomato-based microcosms inoculated on a substrate (apple, banana, cucumber, lemon, tomato). Flies were left to develop in these units and after eclosion they were sexed and dried; development time in days and dry weight in mg was recorded. This experiment was repeated 5 times, with some variation, including I) one experiment where substrate samples were meta-barcoded for microbiota community analysis and II) one experiment after which developed flies and substrate were pooled and subjected to a subsequent Compound Specific Isotope Analysis of amino acids based on δ¹³C values of essential amino acids; the δ¹³C data is also included in the data provided, as well as estimates of fly and substrate amino acid source (plant, bacteria or fungi) based on a Bayesian mixing model (R package, MixSIAR) and a training data set by Larsen et al. (2009; 2016) and Pollierer et al. (2020). For more detailed description of the datasets see Riedel et al. (preprint, https://doi.org/10.1101/2025.10.14.682149) and Riedel and Rohlfs (in preparation).

Funded by the University of Bremen in cooperation with the AG Scheu at the University of Göttingen (Georg-August Universität)