Resistance evolution driven by niche construction – the role of microbiota and beetle secretions

Resistance evolution driven by niche construction – the role of microbiota and beetle secretions | 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 Research Article Resistance evolution driven by niche construction – the role of microbiota and beetle secretions Nora K.E. Schulz, Ana Korsa, Helle Jensen, Lai Ka Lo, Jeanne Friedrichs, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6353988/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Through niche construction, organisms actively shape their environment, thereby influencing their evolutionary trajectories via ecological inheritance. Red flour beetles ( Tribolium castaneum ) achieve niche construction through secretion of antimicrobial compounds from the stink glands. It has recently been demonstrated that the experimental removal of niche construction using RNAi of a key gene needed to produce stink gland secretions altered the pace and mechanisms of resistance adaptation to the bacterial entomopathogen Bacillus thuringiensis within nine host generations. However, it is unknown whether the microbiome and secretions produced by beetles undergo changes during experimental evolution. We continued the evolution experiment with an additional nine generations of selection. We found that host resistance continued to increase in selection regimes with pathogen exposure, whereas host development and fecundity remained stable, thereby confirming our previous findings. We then profiled larvae-associated microbiota in generations 12 and 15 via 16S rRNA sequencing and measured the stink gland secretion profiles of adults via gas chromatography-flame ionization detection in generation 18. While adaptation to the pathogen did not affect the microbiota, lines evolving with the possibility to construct their niches showed increased microbial diversification, and chemical secretion profiles did not change in either of the selection regimes. Together, our results highlight the role of niche construction in shaping host–microbe interactions. These effects seemed to be independent of any microevolutionary changes in the secretions as a niche-constructing trait. Evolutionary Biology niche construction microbiota flour beetles external immunity quinones experimental evolution Figures Figure 1 Figure 2 Figure 3 Introduction Many organisms reshape their environments in ways that influence both their ecological interactions and evolutionary trajectories. This process, known as niche construction, can generate feedback loops that are suggested to impact an organism’s adaptation and survival (Laland, Odling-Smee, and Feldman 1999 ; Lo and Schulz et al. 2025). Such feedback is created by the inheritance of the changed niche from parents to offspring (called ‘ecological inheritance’), which, combined with genetic inheritance, could speed up the adaptation processes. In dense populations, where close contact increases the risk of disease transmission, these environmental modifications may play an important role in mediating or suppressing pathogen infections. For instance, many insects release different volatile and non-volatile chemical compounds to navigate social interactions and defend themselves against external threats. These secretions have different purposes: they can mediate communication and act as a defence against predators and parasites (Yew and Chung 2015). Many tenebrionid beetles have evolved exocrine glands that influence their external immune defences by producing a mixture containing different quinones ( Roth and Howland 1941 ; Joop et al. 2014 ). Besides protection in the form of “external immunity” (Cotter and Kilner 2010 ; Otti, Tragust, and Feldhaar 2014), this emission of defensive secretions into the environment can also represent a niche construction (Odling-Smee, Laland, and Feldman 2013; Müller et al. 2020;). For example, defensive secretions from the stink glands of some beetle species (Cotter and Kilner 2010 ) exhibit antimicrobial properties, thereby modifying the microbial environment in which these beetles reside (Yezerski et al. 2007 ; Li et al. 2013 ). Regarding niche construction, both the defensive secretions in the environment and the microbial community of the environment or the organism itself can be transmitted from parent to offspring (Körner, Steiger, and Shukla 2023 ). Thus, they could represent potentially relevant components that contribute to ecological inheritance. Insects can obtain their microbiota through both environmental sources as well as from parent to offspring (Geib et al. 2009 ). A common form of environmental transfer in insects occurs through the consumption of feces from members of the same species, known as coprophagy. Studies on the red flour beetle ( Tribolium castaneum , Tenebrionidae) have revealed that its microbiota reflects the flour microbiota shared across different life stages, with variations due to beetle population density and grain type, indicating the influence of coprophagy and diet (Agarwal and Agashe 2020; Bi, Zhang, and He 2024). Similar patterns have been observed in other insect species, where microbial transfer via coprophagy enhances host fitness (Jahnes, Herrmann, and Sabree 2019 ; Vera-Ponce de León et al. 2021), suggesting that such behaviours may serve an adaptive function. More broadly, acquisition of commensal microbes, whether through vertical or horizontal transmission, can offer significant benefits to the host, including improved immunity and fitness. For instance, despite variation in the microbiota of T. castaneum across life stages and environmental conditions, it provides immune and fitness advantages and is further shaped by the host’s immune experiences (Futo, Armitage, and Kurtz 2016; Agarwal and Agashe 2020; Korša et al. 2022 ). These observations underscore the significance of microbiome-mediated effects on ecological niches and host immunity, suggesting a complex interplay between genetic and ecological inheritance. The microbiome has also been shown to play a vital role in determining resistance (Zhang et al. 2015 ), with certain gut bacteria influencing the host response to environmental stress and chemical threats. For example, specific bacteria like Enterococcus faecalis can alter the physiological stress responses of T. castaneum , thereby modulating its susceptibility to fumigants (Wang et al. 2023 ). The speed at which resistance develops is greatly influenced by environmental factors, such as climate and the diversity of available habitats (Maino, Umina, and Hoffmann 2018 ). Moreover, the consumption of pathogen-infested food has been shown to alter quinone production levels, affecting the beetle's chemical defences against pathogens (Davyt-Colo et al. 2022). A recent study provided insights in how niche construction affects selection, especially in the context of host-parasite interactions (Lo and Schulz et al. 2025). In this study, the effect of niche construction on resistance evolution and host adaptation to a pathogen was assessed using the experimental evolution of a well-established host-parasite model with T. castaneum and the entomopathogenic bacterium, Bacillus thuringiensis tenebrionis ( Btt ) (Lo and Schulz et al. 2025). Adult red flour beetles modify or construct their flour environment using potent antibacterial secretions from their stink glands (SGS), thereby providing external immune protection for both the adults and their offspring (Cotter et al. 2013 ; Joop et al. 2014 ; Müller et al. 2020). In this experimental evolution setup, experimental removal of niche construction—using RNAi of a key gene needed to produce stink gland secretions—was used to test the effects of the beetles’ adaptation to Btt . In each generation, beetles inhabited conditioned flour either with regular or drastically reduced stink gland secretion content. Importantly, as the RNAi treatment was performed in siblings of the beetles exposed to the niches, also the niche-constructing trait itself–the SGS–was allowed to evolve. After only three generations, populations with niche construction and exposure to the pathogen showed improved survival against pathogen infection compared to populations without niche construction (Lo and Schulz et al. 2025). By the ninth generation, this enhanced survival was apparent in all Btt- selected populations but was particularly strong in those inhabiting constructed niches. Moreover, RNAseq of evolved beetles showed that gene expression differed strongly between the selection regimes, revealing that the mechanisms underlying resistance were dependent on niche construction. It is yet unclear whether the observed differences after nine generations of evolution are stable and will prevail for a longer period of selection, and what causes the observed differences caused by the experimental removal of niche construction. In addition to genetic adaptations achieving resistance–and which might differ in the presence or absence of niche construction–ecological inheritance could have driven the adaptation process. It is therefore, important to gain insights into two potentially relevant components of ecological inheritance: the SGS and the microbiota. In the present study, we continued this experimental evolution experiment and assessed the survival, development, and fitness measures of generations 12, 15, and 18. Furthermore, with 16S rRNA sequencing, we assessed the microbiota composition of the larvae in generations 12 and 15. We aimed to examine whether niche and pathogen selection shape overall stink gland secretion profiles. We thus examined the chemical secretions of the beetles from the different evolution regimes in the 18th generation using gas chromatography-flame ionization detection. Overall, this study sheds light on how organisms actively modify their environment and how these changes, in turn, impact their microbiome and chemical ecology. Materials and methods Niche construction and experimental evolution design Tribolium castaneum beetles used in this study originated from experimentally evolved lines described in Lo and Schulz et al. (2025). Beetles were maintained in glass jars containing heat-sterilized (75 °C) organic wheat flour (type 550, Drogerie-Markt (DM) supplemented with 5% brewer’s yeast (hereafter referred to as “flour”). All lines were kept at 30 °C and 70% relative humidity under a 12-hour light/dark cycle. In each generation, we used siblings from the previous generation as flour-conditioning beetles ( niche constructors ) as described in Lo and Schulz et al. (2025). To produce secretion-free lines, we impaired the proper development of stink glands and the subsequent production of SGS ( niche-free regimes) through pupal knockdown of Drak gene using RNAi ( Drak -dsRNA 1000 ng/µl in phosphate-buffered saline (PBS), Lo and Schulz et al 2025). For simplicity, we call this treatment ‘ niche-free’ . For niche regimes, control beetles received eGFP- dsRNA injections at the same concentration, which does not affect target gene expression. The experimental evolution experiment consisted of four regimes as described in Lo and Schulz et al. (2025; Figure S1): two regime groups, each with four replicate lines, were subjected to pathogen selection after developing in either constructed or non-constructed niches ( niche/pathogen and niche-free/pathogen, respectively), and two regime groups, each with four replicate lines, developed in constructed or non-constructed niches without pathogen selection ( niche/pathogen-free and niche-free/pathogen-free , respectively). To start the selection experiment, freshly laid eggs were placed in jars containing flour that had been conditioned with 100 flour-conditioning beetles for four days to establish the initial generation. On the 15th day post-oviposition, 192 larvae per line were randomly chosen and exposed to either the pathogen infected or the control treatment. Following a four-day exposure period, 100 surviving larvae were selected and transferred to jars containing freshly conditioned flour (0.2 g per beetle). This process of exchanging conditioned flour continued until the selection line individuals emerged as adults and began producing their secretions. After one to two months, these adult beetles were used to generate the subsequent generation of niche constructors and inhabitants. Pathogen infection assays For the selection and survival assays, Bacillus thuringiensis tenebrionis ( Btt ; BGSCID 4AA1) was obtained from the Bacillus Genetic Stock Centre (BGSC, Ohio State University, USA) and the assays were prepared as described in Lo and Schulz et al. (2025). Methods for cultivation and oral infection were adapted from Milutinović et al. (2013). To prepare overnight cultures, Btt from glycerol stock was introduced into 3-4 ml of Bt culture medium, including 15 μl of filter-sterilized salt solution (using 0.2 μm cellulose nitrate filters, Whatman) and 3.75 μl of 1M CaCl 2 . This was incubated at 30 °C and 180 rpm. The next day, 400 mL of Bt medium, supplemented with 1 mL of salt solution and 500 μl of CaCl 2 , was inoculated with the overnight culture and incubated in darkness at 30 °C and 180 rpm for eight days until sporulation was complete. After an eight-day sporulation period, the spores were rinsed and diluted to 2 x 10^10 spores/ml using phosphate-buffered saline (PBS, Calbiochem). This suspension was then combined with a heat-sterilized flour mixture at a ratio of 0.15 g per mL of spore suspension. Under sterile conditions, 10 μL of this mixture was dispensed into each well of flat-bottom 96-well plates (Sarstedt), sealed with a breathable adhesive film (Kisker Biotech GmbH & Co. KG) and left to dry overnight at 30°C in darkness. For the regime groups without Btt exposure ( niche/pathogen-free and niche-free/pathogen-free ), the plates were prepared identically, but without bacteria. Phenotypic readouts in generations 12, 15 and 18 During generations 12, 15, and 18 of continuous experimental evolution focusing on adaptation to pathogens in both niche and niche-free environments, we conducted phenotypic evaluations across all replicate lines following two generations of relaxed selection, referred to as the first (F1) and second (F2) relaxation generations as detailed in Lo and Schulz et al. (2025). We evaluated survival rates post-pathogen infection, the speed of development until pupation, and early-life reproductive capacity. All experimental subjects were derived from two sets of 24-hour egg-laying periods by 100 F1 (from generations 12,15, and 18) adults in glass containers, following previously established protocols. Survival after pathogen infection To assess the development of resistance to Btt , we subjected 96 larvae per line from all lines across the four selection regimes to a concentration of 2 x 10 10 spores/ml of Btt (LD50 in the original population) and monitored their survival over four days (for detailed methods see Lo and Schulz et al., 2025). The oral infection of Btt for the survival test followed the same methodology as the selection protocol. We randomly selected F2 larvae from each selection line at 15 days of age, with 96 individuals per replicate line exposed to Btt spores. Additionally, we randomized larvae from different regimes across different plates to avoid potential plate effects. Developmental assay To assess the impact of niche construction on development time, we monitored the pupation timing of F2 beetles raised under unrestricted feeding conditions without exposure to pathogens (for detailed methods see Lo and Schulz et al., 2025). At 14 days after egg-laying (dpo), we placed 48 larvae per selection line individually into 96-well plates. Each well contained roughly 0.2 g of flour diet. The larvae were maintained under standard rearing conditions for the duration of the experiment. We recorded pupation daily from 19 dpo through 32 dpo. Fecundity To test for potential costs of evolved resistance or adaptation to the constructed niches, we measured fecundity of F2 beetles as a proxy of reproductive fitness (for detailed methods see Lo and Schulz et al., 2025). 24 days after oviposition, the majority of beetles from a single F1 beetle oviposition had reached the pupal stage. We identified their sexes and placed the F2 pupae individually. After all beetles had reached sexual maturity, we placed 20 pairs of one male and one female from each line into vials with 5 g of flour for three days. Following this interval allowing for mating, the adult beetles were removed from the flour. Twelve days later, we counted the number of living larvae in each vial. Statistical analysis of phenotypic readouts All analysis were performed in R (R Core Team 2024; version 4.4.2) using R Studio (RStudio Team 2023; 2023; Posit Software, version 2024.12.0). For survival and development, data we performed Cox proportional hazard analysis using the “survival” package (version 2.38; Therneau et al. 2024; Therneau, Grambsch, and Pankratz 2003) on each generation separately. Hazard ratios, or progression ratios, respectively, were calculated compared to the internal control, the niche/pathogen-free regime, while pooling replicate lines for the analysis. For the fecundity data analysis, we excluded pairs with less than five offspring, as we assume that these occurred rather due to technical reasons (premature death of one parent from handling, misassignment of sex, etc.) than that they were of biological relevance. A GLMM with negative binomial error distribution was fitted to the remaining data using the “lme4” package (Bates et al. 2015). The number of larvae served as response variable, replicate line as random factor and selection regime as explanatory variable with the three regimes compared against the internal control niche/pathogen-free regime. On generation 18, we additionally performed a principal component analysis (PCA) to visualize how the three phenotypic readouts (survival of infection, pupation rate, and fecundity) contribute to the differences between regimes and how they relate to each other. Principal components were calculated with the standardized means for each of the four replicate lines per selection regime using the “prcomp” function and results were visualized using “ggbiplot” function from the ggplot2 package (Wickham et al. 2024). Microbial community profiles To examine the impact of selection regimes on the microbiota of the beetles, we analysed the F2 of evolved beetle lines from generations 12 and 15. We took three biological replicates from each selection line within both generations, resulting in 96 samples. RNA was extracted from pools of larvae (see below) and transcribed to cDNA before performing 16S rRNA amplicon sequencing of the V3-V4 hypervariable region. RNA from biological replicates was extracted on different days to ensure randomization. We then clustered sequences to operational taxonomic units (OTUs). RNA extraction and cDNA synthesis For each replicate, we pooled twenty 14-day-old larvae, transferred them to sterilized 1.5 ml Eppendorf tubes, froze them in liquid nitrogen and stored them at -80 °C. Generation 12 larvae were stored for approximately one year, while generation 15 larvae were extracted within a week. Using a sterile pestle, larvae were ground in liquid nitrogen. Half of the sample was transferred to a new tube with 100 μl phenol-chloroform-isoamyl alcohol and briefly mixed. Subsequent steps followed the RNeasy Power Microbiome Kit (QIAGEN N.V., Netherlands) protocol, with minor alterations. We added 70% ethanol to avoid co-purification of small RNA species. RNA was eluted in 50 μl of nuclease-free water. RNA concentration and quality were measured with a NanoPhotometer (Pearl, Implen, USA), Qubit BR RNA assay (Invitrogen Qubit Fluorometer 2.0, Thermo Fisher Scientific Inc., USA) and a subset with the Agilent 2100 Bioanalyzer for RNA integrity (Agilent Technologies Inc, USA). RNA was stored at -80 °C. For cDNA synthesis, the RevertAid II Kit (Thermo Fisher Scientific Inc., USA) with random hexamer primers was used, adding 5 μl RNA per reaction. Samples were transported to the sequencing facility on dry ice. 16S rRNA sequencing and processing The library preparation and sequencing of 16S rRNA was executed by Novogene (Novogene UK, Cambridge Sequencing Center). Amplification of cDNA was performed with 10 ng template DNA following PCR purification using magnetic beads. The V3-V4 hypervariable region was amplified (primers: CCTAYGGGRBGCASCAG; GGACTACNNGGGTATCTAAT) and sequenced on a paired-end Illumina platform to generate 250bp paired-end raw reads. Splitting of paired-end reads and truncation of barcodes and primers was performed using Python (V3.6.13) and cutadapt (V3.3). Using FLASH (V1.2.11) and fastp (V0.23.1) the paired-end reads were merged and filtered to obtain high-quality merged sequences. Chimeric sequences were removed with vsearch (V2.16.0). Operational taxonomic units (OTUs) were clustered using Uparse software (V7.01001), with sequences ≥ 97% similar assigned to the same OTUs. Species annotation used the Mothur algorithm applied to the SILVA 138.1 database (Quast et al. 2013). The SILVA reference was filtered to cluster 99% similar sequences for accurate taxonomy assignments. As SILVA only provides species-level taxonomy, NCBI searches supplemented information on other taxonomic levels. MUSCLE software (V3.8.31) constructed phylogenetic relationships through multiple sequence alignment. Bioperl's (V1.7) SVG function constructed the phylogenetic tree. Analysis of microbiota data R (V4.3.3, R Core Team 2024) and R studio (V 2023.06.0 + 421, RStudio Team 2023) were used to conduct all data visualization and analysis. Taxonomy, sample information and OTU count data were merged using the package “phyloseq” in R (McMurdie and Holmes 2013) with the help of the packages “biomformat” (McMurdie & Paulson 2017), “ape” (Paradis and Schliep 2019) and “Matrix“ (Bates, Maechler, and Jagan 2000). Low abundant taxa were filtered out by removing OTUs that occurred with less than ten copies in individual samples, as described in (Nikodemova et al. 2023). In the next step, Cyanobacteria, Archaea and unassigned OTUs were removed using the functions available in “phyloseq”. Alpha diversity was assessed using the Chao1 richness estimator to evaluate differences in microbial richness across experimental regimes. Chao1 values were calculated using the “estimate_richness()” function from the “phyloseq” package. To compare richness across regimes, a Kruskal-Wallis rank sum test was applied, as the data did not meet the assumptions of parametric testing. Beta diversity was assessed using Aitchison distance (Aitchison et al. 2000), which accounts for the compositional nature of microbiome data, using the “vegan” package (Oksanen et al. 2001). The OTU table was CLR-transformed using the “clr()” function from the “compositions” R package after applying a pseudocount of 1 to avoid issues with zero counts. PCA was performed on the clr-transformed data to visualize differences in microbial community composition across regimes and generations. To test for differences in community composition between regimes, we used PERMANOVA (“adonis2” from the “vegan” package) with 999 permutations. Assumptions of homogeneity of group dispersions were assessed using “betadisper()” followed by an ANOVA. Pairwise comparisons between regimes were conducted using the “pairwise.adonis” function with Bonferroni-adjusted p-values. Furthermore, ANOSIM was performed to complement PERMANOVA results and evaluate the relative strength of group separation. We applied Random Forest classification to identify microbial taxa (OTUs) most important for distinguishing microbiota between experimental regimes. The OTU table was CLR-transformed using the “compositions” package after adding a pseudocount of 1 to account for zero inflation. Random Forest models were trained using the “randomForest” package (Breiman et al. 2002) in R with 1000 trees and default parameters. Feature importance was assessed using the Mean Decrease in Accuracy metric. The top 20 OTUs were visualized in a horizontal bar plot, with taxonomy assigned from the phyloseq object and colored by the regime with which they were most associated. Taxonomic labels were derived at the genus level when available. Stink gland secretion profiles Extraction method and data compilation To extract stink gland secretions, groups of six beetles of the same sex were transferred to pre-cooled 2 ml microreaction tubes and placed in ice water for 3 min to induce excretion of gland secretion as described before (n = 5 samples per treatment group and sex; Joop et al. 2014; Lo et al. 2023). Beetles inside the tubes were freeze-killed directly after secretion at -20°C. For the extraction, we added 360 µl ice-cold acetone (≥ 99.8% HPLC grade, Fisher Scientific, Loughborough, UK) containing 0.05 mg/ml n -octadecane (≥ 98.5%, Sigma-Aldrich, St. Louis, USA) as internal standard. Samples were shaken for 5 min at 4°C, centrifuged and supernatants were transferred to vials. Replicates of the different treatments were prepared fresh and either directly analysed or stored for one day at -20°C. In addition, at each extraction day, one blank without beetles was prepared (total n = 5). Beetle samples and blanks were analysed using a gas-chromatograph connected to a flame ionization detector (GC-2010 Plus-FID, Shimadzu, Kyoto, Japan) equipped with a VF-5ms column (30 m × 0.25 mm × 0.25 μm, 10 m guard column, Varian, Agilent Technologies, Santa Clara, California, USA), with nitrogen as carrier gas. The samples were injected at 250°C at a split of 5 with a constant column flow rate of 1 ml/min. The temperature program started at 50°C for 2 min, was then increased in 10°C/min steps to 300°C and the final temperature was maintained for 10 min. Alkane standard mixtures C7-40 and C21-C40 (both Sigma-Aldrich, St. Louis, USA) were analysed using the same temperature program to calculate retention indices (Kováts 1958; van Den Dool and Dec. Kratz 1963). In addition, standards (30 ng/µl) of 2-methyl-1,4-benzoquinone (MBQ), 2-methyl-1,4-hydroquinone (MHQ), 1-pentadecene (all 98%, Sigma-Aldrich) and 2-ethyl-1,4-hydroquinone (EHQ, 97%, Abcr GmbH, Karlsruhe, Germany) were measured for metabolite verification. Pre-processing of the data such as peak integration was done in GCSolution Postrun analysis (Version 2.30, update 2.41, Shimadzu, Kyoto, Japan), with the following parameters: width of 3 sec., slope of 1200 uV/min, min. area/height of 0 counts and default band time of 0.03 min. Metabolites of interest were extracted automatically or integrated manually. Alignment of the retention times was performed in R (version 4.4.1, (R Core Team 2018)) and RStudio (version 2024.04.2+764, R Studio Team 2024) using the package “GCalignR” (Ottensmann et al. 2018) with a maximum linear retention time shift (max_linear_shift) of 0.03 min. Peak areas of the metabolites were normalized to the area of the internal standard. Only metabolites were further considered, whose areas were at least five times higher than the averaged areas in the blank samples. To identify metabolites, eight samples of the different treatment groups and sexes, the two alkane series and the standards were analysed using a GC coupled to a mass spectrometer (GC 2010 Plus – MS QP2020, Shimadzu, Kyoto, Japan) in electron impact ionization mode with an identical column as for the GC-FID and helium as carrier gas, but a split of 8. The temperature program was the same as used for the GC-FID measurements. The ion source temperature was set to 230°C and the interface temperature to 250°C. In quadrupole MS mode at 70 eV the line spectra were acquired with m / z 40 to 600 and a scan event time of 0.25 sec. The mass spectra and retention indices were compared to those in the NIST database (NIST14, National Institute of Standards and Technology, Gaithersburg, Maryland) and Pherobas (El-Sayed 2024) for the putative identification. Metabolites measured by GC-FID were identified based on the RI. Data analysis of secretion profiles All data analysis and visualization were conducted using R (V4.3.3, R Core Team 2024) and R studio (V 2023.06.0 + 421, RStudio Team 2023). To test for differences in compound amounts—represented by the normalized area of the GC peak—we focused on five major representatives of stink gland secretions (SGS), the four quinones (EBQ, EHQ, MBQ, MHQ) and the carbohydrate 1-pentadecene. For each sex and metabolite separately, we fitted a linear mixed-effects model to the log-transformed data using the “lme4” package (Bates et al. 2015). The technical GC run was included as a random effect to account for variability across analytical replicates. Pairwise comparisons of model estimates were performed using the “emmeans” package (Lenth 2017). To explore potential differences in the multivariate composition of SGS profiles among selection regimes, we first performed non-metric multidimensional scaling (NMDS) using a Bray–Curtis dissimilarity matrix, implemented in the “vegan” package (Oksanen et al. 2001). NMDS was conducted separately for each sex. This unsupervised ordination method provides a visual representation of sample relationships while making minimal distributional assumptions. As the NMDS plots did not reveal any group separation by regime, we followed up with a linear discriminant analysis (LDA) to formally assess whether group-level differences could be detected in a supervised classification framework. LDA was conducted using the “MASS” package (Ripley and Venables 2009) and scores for the first two linear discriminants (LD1 and LD2) were visualized using “ggplot2” (Wickham et al. 2024). Finally, we combined data across sexes and ran additional LDAs using only “niche” and “pathogen” as treatment factors, respectively. The resulting LD1 scores were visualized as density distributions to evaluate treatment-level separation. Results Enhanced pathogen resistance continues to evolve without apparent costs We continued an evolution experiment testing the influence of niche construction (i.e. the presence or absence of defensive secretions) on pathogen adaptation and added another nine generations of experimental evolution (Figure S1, Lo and Schulz et. al 2025). Four replicate beetle lines had evolved with impaired vs. normal ability to construct their niches (for simplicity called ‘ niche-free’ vs. ‘ niche’ ) and with vs. without Btt exposure (called ‘ pathogen’ vs. ‘ pathogen-free’ ). To test for continued evolution of resistance we measured survival of Btt infection in the F2 offspring every third generation. The niche/pathogen-free regime was considered the control, because it resembles the unselected situation, where beetles can construct their niche and do not encounter any pathogens. In the regimes with pathogen selection, host resistance continued to increase during the evolution experiment (Figure 1A; Table S1). After a total of 18 generations of selection, 92% ( niche/pathogen ) and 94% ( niche-free/pathogen ), respectively, survived the Btt exposure in the pathogen selected regimes, while in the control ( niche/pathogen-free ) the survival was only 57%. These survival differences demonstrate a significant decrease in hazard ratios, with the pathogen selection reducing mortality probabilities to 0.15 ( niche/pathogen ; CI= 0.099-0.217; p<0.001; Figure 1A; Table S1) and 0.11 ( niche-free/pathogen ; CI=0.072-0.172; p<0.001) compared to the niche/pathogen-free regime. Across three time points (generations 12, 15, and 18), F2 offspring from the niche-free/pathogen-free regime exhibited lower mortality than those from the niche/pathogen-free control (Figure 1A; Table S1). Hazard ratios ranged from 0.87 to 0.72 (Table S1), with significantly lower values in generations 12 and 18 (p<0.05; Table S1). This indicates that resistance can emerge even without Btt selection pressure when beetles are prevented from constructing their niche. In contrast to survival, we did not find any consistent changes in development over the course of the experiment (Figure 1B). In generation 12, all three regimes had a transiently faster development to pupation compared to the niche/pathogen-free control, with the progression ratio (PR, i.e. probability to pupate over time relative to control) being more than twice as high in the pathogen-selected regimes ( niche/pathogen : PR=2.01, p<0.001; niche-free / pathogen : PR=2.46, p<0.001). The larvae within the niche-free/pathogen-free regime also pupated significantly earlier (PR: 1.57, p<0.001). However, this faster development disappeared in generation 15 with all four regimes pupating at a similar time (Figure 1B; Table S1). In generation 18, finally, only individuals of the niche/pathogen (PR = 1.49; p<0.001) and niche-free/pathogen-free regime (PR = 1.16; p<0.05) developed significantly faster than individuals of the niche/pathogen-free control. Similarly, for fecundity we did not observe any consistent changes over the course of the evolution experiment (Figure 1C). Only in generation 18 did the pathogen-selected regimes show a decrease in reproduction, with 18% ( niche/pathogen ) and 9% less offspring ( niche-free/pathogen ; Table S1) hinting at possible costs of the evolved resistance. Finally, we performed a principal component analysis (PCA) combining the three phenotypic readouts from generation 18 (Figure 1D). Principal Component 1 (PC1), accounting for over 50% of the variation, was primarily driven by mortality and fecundity differences and clearly separated the regimes based on pathogen selection. Principal Component 2 (PC2) explained 33% of the variation, mainly reflecting variation in development time, but did not align with any specific selection regime. The microbial community differs between some regimes To determine whether Btt selection and evolved resistance in combination with differentially conditioned niches affect the microbiome composition and diversity, we examined the V3-V4 region of bacterial 16S rRNA. We collected larvae from the F2 beetles of generation 12 and generation 15. For this, OTUs were clustered from raw Illumina Amplicon sequencing reads. After filtering, 17,794,626 reads remained with an average of 183,450 reads per sample across 95 samples. Of 6,875 OTUs in the original dataset, 4,487 OTUs remained after filtering. Of those remaining OTUs, only 11.03% could be assigned at the species level, 48.2 % at the genus level, and 74.4% at the family level. The most common genera were largely consistent between generations, with some fluctuations in their abundance (Figure 2A). However, since it has been previously shown that the microbiota of T. castaneum is dependent on its environment (Agarwal and Agashe 2020), fluctuations between generations were expected. To focus on regime effects and potential selection-driven changes in the microbiota, we analysed the generations jointly. Microbial richness, as measured by the Chao1 index did not differ significantly across regimes (Figure 2B; Kruskal-Wallis χ² = 4.40, df = 3, p = 0.221). A PCA of Aitchison distance revealed weak clustering by regime, with some overlap between niche and pathogen treatments (Figure 2C). PERMANOVA indicated a statistically significant effect of regime on microbial community composition (R² = 0.095, p = 0.001); however, a significant result from the PERMDISP test (F = 11.02, p < 0.001) suggested differences in within-group dispersion, meaning that although compositional differences exist between regimes, variation in group dispersion may partially contribute to the observed significance. Post hoc Tukey's HSD tests revealed that niche-free regimes had significantly lower dispersion and different microbial composition compared to niche regimes, while pathogen-selected regimes did not differ in composition and dispersion within niche or niche-free regime (Figure 2C; Table S2). ANOSIM also detected a significant but weak separation in microbiota between regimes (R = 0.10, p = 0.001), consistent with high variability in community structure. Random Forest classification was used to identify the most important OTUs contributing to differences in microbial community composition across regimes (Figure 2D). The model was trained on CLR-transformed OTU data using 1000 trees and achieved an overall out-of-bag (OOB) error rate of 45.26%. Classification accuracy varied across regimes, with higher accuracy observed for microbiota of the niche/pathogen and niche/pathogen-free regimes and lower performance for those of the niche/pathogen-free individuals, which exhibited frequent misclassification. The top 20 OTUs were selected based on Mean Decrease in Accuracy and visualized with genus-level taxonomic annotations. These taxa, representing diverse bacterial groups, were identified as key contributors to the observed variation between regimes, highlighting specific microbial signatures associated with each selection treatment. The selection regime had only minor effects on stink gland secretions In the stink gland secretions collected from adult F2 beetles of generation 18, we detected 26 compounds and identified 22 of them, which produced a measurable signal in at least 25% of the samples, including the four quinones of interest—ethylbenzoquinone (EBQ), methylbenzoquinone (MBQ), ethylhydroquinone (EHQ), and methylhydroquinone (MHQ)—as well as the carrier compound 1-pentadecene (Table S3). Because SGS production is known to differ between sexes in T. castaneum (Lo et al. 2023), we analysed the five focal compounds separately for females and males (Figure S2). Incorporating block (GC run) as a random effect due to considerable run-to-run variation, we found no significant differences in amounts of any of the five compounds between selection regimes for either sex (Figure S2, Table S4). This suggests that experimental evolution did not lead to measurable changes in these characteristic SGS compounds. To test whether selection regimes had instead driven changes in SGS composition, we applied NMDS using Bray–Curtis dissimilarity matrices to the male and female datasets including all 26 compounds. In both cases, the 95% confidence ellipses around group centroids overlapped completely in the NMDS space (Figure S3), indicating no clear separation among selection regimes. As a follow-up, we conducted LDA. However, no consistent separation among selection regimes was observed in either the female (Figure 3A) or male (Figure 3B) LDA results. Finally, we combined the datasets across sexes and performed two additional LDAs using either “niche” (Figure 3C) or “pathogen” (Figure 3D) as the treatment factor. Again, no distinct group separation was detected along the first linear discriminant in either case. Discussion Recent findings demonstrated that niche construction can facilitate rapid adaptation to pathogen pressure (Lo and Schulz et al. 2025). Building on this, we continued to experimentally evolve beetle populations under different regimes—either with or without the possibility to construct a niche, and with or without exposure to a pathogen—to investigate whether evolution led to changes in resistance, larval microbiomes or adult secretion composition. Host resistance continued to evolve towards very high levels of resistance in regimes with pathogen exposure. We observed differences in microbiota profiles of larvae from the niche versus niche-free regimes, suggesting that niche construction fosters greater microbial variation. In contrast, chemical secretion profiles of adults showed no significant changes across regimes, indicating that these secretions remained stable throughout the selection process. Adaptation to the pathogen increased both with and without the constructed niche during generations 12, 15, and 18 in T. castaneum . This evolved resistance, which resulted in a greater survival after infection compared to niche/pathogen-free regime, aligns with findings from earlier experimental evolution studies involving the same pathogen but other host species (Masri et al. 2015 ; Ferro et al. 2019 ). The previous study of the same system (Lo and Schulz et al. 2025) used transcriptomic analyses after nine generations of experimental evolution to reveal that similar levels of resistance can arise through distinct mechanisms. Beetles that evolved pathogen resistance without a constructed niche exhibited constitutive changes in gene expression, even in the absence of pathogen exposure, suggesting a form of general, always-active resistance. In contrast, beetles that evolved in the presence of a constructed niche showed inducible expression changes, activated only upon Btt exposure. The continued evolution in both regimes that we now observed suggests that both evolutionary trajectories can lead to further increase in resistance. While gaining resistance to pathogens often involves trade-offs in life history traits, as found in other insect species (Cotter and Kilner 2010 ; Schwenke, Lazzaro, and Wolfner 2016), both pathogen-selected regimes exhibited nearly complete resistance with seemingly minimal associated costs in the present study. However, it is important to note that in generation 18, the individuals of the niche/pathogen regime generated approximately 9–18% fewer offspring than those of the niche-free/pathogen-free and niche/pathogen free regimes (Fig. 1 C). Under natural conditions, even if low, such a reproductive cost would maybe result in non-resistant lines outcompeting and eventually replacing resistant ones in the absence of continued pathogen pressure. Additionally, there could be some hidden costs that we did not capture in our readouts, such as larval size which can be important for the infection outcome (Barber 2005 ; Horn, Liang, and Luong 2023). We observed substantial variation in developmental time across generations, but without a consistent trend. The individual variation that we found may be driven by some uncontrolled factors such as subtle differences in parental age that we did not control for. Some variation between generations was also observed in the microbiota analysis. However, fluctuations in microbial diversity is expected as T. castaneum microbiota depends on its environment (Agarwal and Agashe 2020). Our results showed that, although overall microbial richness (Chao1 index) was relatively stable under niche and pathogen manipulations, microbial communities in niche-free regimes seemed to be more consistent, with less variability than niche regimes. These findings suggest that constructed niches such as flour of T.castaneum can favour greater variation in microbial community compared to non-constructed niches regardless of infection status. The presence of quinones, which are known to inhibit microbial growth (Armstrong, Spink, and Kahnke 1943; Duarte et al. 2022 ), may have contributed to the increased compositional variation of microbiota observed in larvae of the niche regimes. By selectively favoring quinone-resistant or less susceptible microbial taxa, the ability to release quinones for niche construction may support greater microbiome diversity. This is consistent with previous findings suggesting that quinones can enhance pathogen defense while sustaining beneficial microbial populations (Yezerski et al. 2007 ; Sawada et al. 2020 ). Interestingly, several potentially pathogenic genera—such as Bacillus , Staphylococcus , and Pseudomonas —appeared to be more prevalent in niche-free regimes (Fig. 2 D). Moreover, the specific microorganism taxa influenced by the quinones might differ significantly, as it is known that microbes exhibit varying levels of susceptibility to quinones (Yezerski et al. 2007 ; Duarte et al. 2022 ). This variability highlights the complexity of microbial interactions and the potential selective pressures exerted by quinones. Conversely, in individuals from the non-constructed niches regimes where the environment was less controlled, there tended to be a lower variation in larval microbiota, illustrating how environmental control and niche construction can influence microbial community structure. Host-parasite interactions can influence and be influenced by the microbiome of the host, leading to changes in abundance of specific bacterial taxa (Fredensborg et al. 2020 ). Unlike in the western corn rootworm, Diabrotica virgifera (Chrysomelidae), where microbial diversity was found to be significantly decreased in resistant animals (Paddock et al. 2021 ), in our study, microbial variation was reduced only in larvae of niche-free regimes. While we observed variation in microbiota composition across evolutionary regimes, this variation did not seem to impair host fitness of T. castaneum , indicating that functional stability within the microbial community is preserved (Fassarella et al. 2021 ). By altering their microbial environment through stink gland secretions, beetles may reduce the risk of infection by eliminating pathogens, thereby enhancing their survival and reproductive success (Joop et al. 2014 ). In the experimental evolution over 18 generations, pathogen selection led to strong adaptations, with the beetles almost completely losing their susceptibility to the Btt infection. Via transcriptome analysis, Lo & Schulz et al. (2025) observed clear differences in resistance mechanisms depending on niche construction already after nine generations. While in the presence of a constructed niche the pathogen resistance was achieved mainly through changes in inducible responses, in the regime without the constructed niche resistance correlated with changes in constitutive responses (Lo and Schulz et al. 2025). Does this mean that the external immune response also diverged at a similar rate between the selection regimes? Does the repeated development in conditions with strongly reduced SGS lead to a higher output of this external immune response in adulthood? In other words, can we observe a response to the constant eco-evo feedback created by manipulating the constructed niches? At least in regard to the amount and composition of secreted metabolites the answer appears to be “no”. We found neither significant differences between the regimes when comparing the amounts of the most abundant metabolites in the secretions, nor was there any distinct differentiation in the metabolite composition of the SGS. While our employed method has its limits in sensitivity, the large sample size we used should have enabled us to also detect minute but consistent differences in secretion composition between the regimes. Thus, it appears that the SGS metabolites and the external immunity they provide are not part of the toolbox involved in resistance evolution. This might be explained by the fact that Btt infects internally via the midgut after ingestion (Palma et al. 2014 ). The spores only germinate inside the host (Raymond et al. 2010 ), thus more sensitive vegetative cells do not come into direct contact with the host’s secretions. Therefore, the adaptation via more Btt specific immune traits, e.g. upregulation of specific antimicrobial peptide gene expression in the midgut (Baur et al. 2024 ) is likely to be more successful. However, we only analyse changes in metabolites here, but the beetles could also alter their external immune functions by altering the protein composition of the secretions. To test this, proteomic analysis of the SGS or detecting changes in antimicrobial activity via zone of inhibition assays (Lo and Schulz et al. 2025) would be useful as well as detailed analysis of the conditioned flour niche. Finally, we also did not find an effect of evolving in a constructed niche on the SGS profiles, thus the beetles also responded to the less controlled microbial niche caused by the drastically lowered SGSs with altering other traits than their quinone defence. This might also help us to understand the relatively small changes in their associated microbiota. It is also important to note that both SGS profiles and microbiome composition were analysed in F2 offspring. This supports the idea that the observed differences between selection regimes are stable and likely reflect underlying genetic changes rather than transient environmental effects. If these differences are indeed genetically encoded, it is reasonable to assume that similar microbiome patterns would be detectable across different life stages, including both larvae and adults. Overall, this study provides novel insights into the modulation of host-associated microbiota within complex ecological niches. Our data indicates that a niche constructing trait itself, the SGS, did not change during evolutionary adaptation to a pathogen in the presence vs. absence of a constructed niche, while the larval microbiota showed small but potentially relevant differences between these niche regimes. However, more studies are needed (e.g., on the niche associated microbiome) to conclude whether this has significantly contributed to ecological inheritance. Declarations Competing interest The authors declare no competing interests. Author contribution N.K.E.S. and J.K. conceptualized the study. N.K.E.S, H.J., J.F., C.M., and J.K. designed the experiments. N.K.E.S. and L.K.L maintained the selection lines. N.K.E.S., H.J., and J.F. performed the experiments. N.K.E.S., A.K. and J.F. analysed the data. N.K.E.S., A.K., and J.F. wrote the original manuscript draft. All the authors revised the manuscript. Acknowledgment We would like to thank Kathrin Brüggemann for all the help in the lab, and Paula Mayer, Tobias Prüser and Tobias Gryczan for the help with selection line maintenance. We acknowledge funding by the Deutsche Forschungsgemeinschaft (DFG, German Research 617 Foundation) as part of the SFB TRR 212 (NC³) – Project numbers 316099922, 396780003 (to JK) and 396782989 (to CM). References Agarwal, Aparna, and Deepa Agashe. 2020. “The Red Flour Beetle Tribolium Castaneum: A Model for Host-Microbiome Interactions.” Edited by Jeffrey P. Demuth. PLOS ONE 15 (10): e0239051. https://doi.org/10.1371/journal.pone.0239051. Aitchison, J., C. Barceló-Vidal, J. A. Martín-Fernández, and V. Pawlowsky-Glahn. 2000. “Logratio Analysis and Compositional Distance.” Mathematical Geology 32 (3): 271–75. https://doi.org/10.1023/A:1007529726302. Armstrong, W. D., Wesley W. Spink, and Jeanne Kahnke. 1943. “Antibacterial Effects of Quinones.∗.” Proceedings of the Society for Experimental Biology and Medicine 53 (2): 230–34. https://doi.org/10.3181/00379727-53-14262. Barber, Iain. 2005. “Parasites Grow Larger in Faster Growing Fish Hosts.” International Journal for Parasitology 35 (2): 137–43. https://doi.org/10.1016/j.ijpara.2004.11.010. Bates, Douglas, Martin Mächler, Ben Bolker, and Steve Walker. 2015. “Fitting Linear Mixed-Effects Models Using Lme4.” Journal of Statistical Software 67 (October):1–48. https://doi.org/10.18637/jss.v067.i01. Bates, Douglas, Martin Maechler, and Mikael Jagan. 2000. “Matrix: Sparse and Dense Matrix Classes and Methods.” https://doi.org/10.32614/CRAN.package.Matrix. Baur, Moritz, Nora K. E. Schulz, Lilo Greune, Zoe M. Länger, Jürgen Eirich, Iris Finkemeier, Robert Peuß, Petra Dersch, and Joachim Kurtz. 2024. “Immune Priming in the Insect Gut: A Dynamic Response Revealed by Ultrastructural and Transcriptomic Changes.” bioRxiv. https://doi.org/10.1101/2024.11.21.624623. Bi, Jie, Lirui Zhang, and Yanping He. 2024. “Metabolomics and Microbiome Provide New Insights into Mechanisms of the Variations between the Larva and Adult Stages of the Model Beetle Tribolium Castaneum (Herbst).” Journal of Asia-Pacific Entomology 27 (2): 102241. https://doi.org/10.1016/j.aspen.2024.102241. Breiman, Leo, Adele Cutler, Andy Liaw, and Matthew Wiener. 2002. “randomForest: Breiman and Cutlers Random Forests for Classification and Regression.” https://doi.org/10.32614/CRAN.package.randomForest. Cotter, Sheena C., and R.M. Kilner. 2010. “Personal Immunity versus Social Immunity.” Behavioral Ecology 21 (4): 663–68. https://doi.org/10.1093/beheco/arq070. Cotter, Sheena C., Joanne E. Littlefair, Peter J. Grantham, and Rebecca M. Kilner. 2013. “A Direct Physiological Trade-off between Personal and Social Immunity.” Journal of Animal Ecology 82 (4): 846–53. https://doi.org/10.1111/1365-2656.12047. Davyt-Colo, Belén, Juan R. Girotti, Andrés González, and Nicolás Pedrini. 2022. “Secretion and Detection of Defensive Compounds by the Red Flour Beetle Tribolium Castaneum Interacting with the Insect Pathogenic Fungus Beauveria Bassiana.” Pathogens (Basel, Switzerland) 11 (5): 487. https://doi.org/10.3390/pathogens11050487. Den Dool, H. van, and P. Dec. Kratz. 1963. “A Generalization of the Retention Index System Including Linear Temperature Programmed Gas—Liquid Partition Chromatography.” Journal of Chromatography A 11 (January):463–71. https://doi.org/10.1016/S0021-9673(01)80947-X. Duarte, Sónia, Ana Magro, Joanna Tomás, Carolina Hilário, Ricardo Boavida Ferreira, and Maria Otília Carvalho. 2022. “Antifungal Activity of Benzoquinones Produced by Tribolium Castaneum in Maize-Associated Fungi.” Insects 13 (10): 868. https://doi.org/10.3390/insects13100868. El-Sayed, Ashraf M. 2024. “The Pherobase: Database of Pheromones and Semiochemicals | The World Largest Database of Behavioural Modifying Chemicals.” 2024. https://pherobase.com/. Fassarella, Marina, Ellen E. Blaak, John Penders, Arjen Nauta, Hauke Smidt, and Erwin G. Zoetendal. 2021. “Gut Microbiome Stability and Resilience: Elucidating the Response to Perturbations in Order to Modulate Gut Health.” Gut 70 (3): 595–605. https://doi.org/10.1136/gutjnl-2020-321747. Ferro, Kevin, Robert Peuß, Wentao Yang, Philip Rosenstiel, Hinrich Schulenburg, and Joachim Kurtz. 2019. “Experimental Evolution of Immunological Specificity.” Proceedings of the National Academy of Sciences of the United States of America 116 (41): 20598–604. https://doi.org/10.1073/pnas.1904828116. Fredensborg, Brian L., Inga Fossdal í Kálvalíð, Thor B. Johannesen, C. Rune Stensvold, Henrik V. Nielsen, and Christian M. O. Kapel. 2020. “Parasites Modulate the Gut-Microbiome in Insects: A Proof-of-Concept Study.” Edited by Simon Russell Clegg. PLOS ONE 15 (1): e0227561. https://doi.org/10.1371/journal.pone.0227561. Futo, Momir, Sophie A O Armitage, and Joachim Kurtz. 2016. “Microbiota Plays a Role in Oral Immune Priming in Tribolium Castaneum.” Frontiers in Microbiology . https://doi.org/10.3389/fmicb.2015.01383. Geib, Scott M., Maria del Mar Jimenez-Gasco, John E. Carlson, Ming Tien, Randa Jabbour, and Kelli Hoover. 2009. “Microbial Community Profiling to Investigate Transmission of Bacteria between Life Stages of the Wood-Boring Beetle, Anoplophora Glabripennis.” Microbial Ecology 58 (1): 199–211. https://doi.org/10.1007/s00248-009-9501-4. Horn, Collin J., Caroline Liang, and Lien T. Luong. 2023. “Parasite Preferences for Large Host Body Size Can Drive Overdispersion in a Fly-Mite Association.” International Journal for Parasitology 53 (7): 327–32. https://doi.org/10.1016/j.ijpara.2023.03.003. Jahnes, Benjamin C., Madeline Herrmann, and Zakee L. Sabree. 2019. “Conspecific Coprophagy Stimulates Normal Development in a Germ-Free Model Invertebrate.” PeerJ 7:e6914. https://doi.org/10.7717/peerj.6914. Joop, G., O. Roth, P. Schmid-Hempel, and J. Kurtz. 2014. “Experimental Evolution of External Immune Defences in the Red Flour Beetle.” Journal of Evolutionary Biology 27 (8): 1562–71. https://doi.org/10.1111/jeb.12406. Körner, Maximilian, Sandra Steiger, and Shantanu P. Shukla. 2023. “Microbial Management as a Driver of Parental Care and Family Aggregations in Carrion Feeding Insects.” Frontiers in Ecology and Evolution 11 (August). https://doi.org/10.3389/fevo.2023.1252876. Korša, Ana, Lai Ka Lo, Shrey Gandhi, Corinna Bang, and Joachim Kurtz. 2022. “Oral Immune Priming Treatment Alters Microbiome Composition in the Red Flour Beetle Tribolium Castaneum.” Frontiers in Microbiology 13. https://www.frontiersin.org/articles/10.3389/fmicb.2022.793143. Kováts, E. 1958. “Gas-Chromatographische Charakterisierung Organischer Verbindungen. Teil 1: Retentionsindices Aliphatischer Halogenide, Alkohole, Aldehyde Und Ketone.” Helvetica Chimica Acta 41 (7): 1915–32. https://doi.org/10.1002/hlca.19580410703. Laland, K. N., F. J. Odling-Smee, and M. W. Feldman. 1999. “Evolutionary Consequences of Niche Construction and Their Implications for Ecology.” Proceedings of the National Academy of Sciences 96 (18): 10242–47. https://doi.org/10.1073/pnas.96.18.10242. Lenth, Russell V. 2017. “Emmeans: Estimated Marginal Means, Aka Least-Squares Means.” https://doi.org/10.32614/CRAN.package.emmeans. Li, Jianwei, Sabrina Lehmann, Bernhard Weißbecker, Irene Ojeda Naharros, Stefan Schütz, Gerrit Joop, and Ernst A. Wimmer. 2013. “Odoriferous Defensive Stink Gland Transcriptome to Identify Novel Genes Necessary for Quinone Synthesis in the Red Flour Beetle, Tribolium Castaneum.” PLOS Genetics 9 (7): e1003596. https://doi.org/10.1371/journal.pgen.1003596. Lo, Lai Ka, Reshma R, Lisa Johanna Tewes, Barbara Milutinović, Caroline Müller, and Joachim Kurtz. 2023. “Immune Stimulation via Wounding Alters Chemical Profiles of Adult Tribolium Castaneum.” Journal of Chemical Ecology 49 (1): 46–58. https://doi.org/10.1007/s10886-022-01395-x. Lo, Lai Ka, Nora K.E. Schulz, Helle Jensen and Joachim Kurtz. 2025 " Experimental Removal of Niche construction Alters the Pace and Mechanisms of Resistance Evolution. " bioRxiv. 2025.03.27.645637; doi: https://doi.org/10.1101/2025.03.27.645637 Maino, James L., Paul A. Umina, and Ary A. Hoffmann. 2018. “Climate Contributes to the Evolution of Pesticide Resistance.” Global Ecology and Biogeography 27 (2): 223–32. https://doi.org/10.1111/geb.12692. Masri, Leila, Antoine Branca, Anna E. Sheppard, Andrei Papkou, David Laehnemann, Patrick S. Guenther, Swantje Prahl, et al. 2015. “Host–Pathogen Coevolution: The Selective Advantage of Bacillus Thuringiensis Virulence and Its Cry Toxin Genes.” PLoS Biology . https://doi.org/10.1371/journal.pbio.1002169. McMurdie, Paul J., and Susan Holmes. 2013. “Phyloseq: An R Package for Reproducible Interactive Analysis and Graphics of Microbiome Census Data.” Edited by Michael Watson. PLoS ONE 8 (4): e61217. https://doi.org/10.1371/journal.pone.0061217. Müller, Caroline, Barbara A. Caspers, Jürgen Gadau, and Sylvia Kaiser. 2020. “The Power of Infochemicals in Mediating Individualized Niches.” Trends in Ecology & Evolution 35 (11): 981–89. https://doi.org/10.1016/j.tree.2020.07.001. Nikodemova, Maria, Elizabeth A. Holzhausen, Courtney L. Deblois, Jodi H. Barnet, Paul E. Peppard, Garret Suen, and Kristen M. Malecki. 2023. “The Effect of Low-Abundance OTU Filtering Methods on the Reliability and Variability of Microbial Composition Assessed by 16S rRNA Amplicon Sequencing.” Frontiers in Cellular and Infection Microbiology 13 (June). https://doi.org/10.3389/fcimb.2023.1165295. Odling-Smee, F. John, Kevin N. Laland, and Marcus W. Feldman. 2013. Niche Construction: The Neglected Process in Evolution . Princeton University Press. https://doi.org/10.1515/9781400847266. Oksanen, Jari, Gavin L. Simpson, F. Guillaume Blanchet, Roeland Kindt, Pierre Legendre, Peter R. Minchin, R.B. O’Hara, et al. 2001. “Vegan: Community Ecology Package.” https://doi.org/10.32614/CRAN.package.vegan. Ottensmann, Meinolf, Martin A. Stoffel, Hazel J. Nichols, and Joseph I. Hoffman. 2018. “GCalignR: An R Package for Aligning Gas-Chromatography Data for Ecological and Evolutionary Studies.” PLOS ONE 13 (6): e0198311. https://doi.org/10.1371/journal.pone.0198311. Otti, Oliver, Simon Tragust, and Heike Feldhaar. 2014. “Unifying External and Internal Immune Defences.” Trends in Ecology & Evolution 29 (11): 625–34. https://doi.org/10.1016/j.tree.2014.09.002. Paddock, Kyle J., Adriano E. Pereira, Deborah L. Finke, Aaron C. Ericsson, Bruce E. Hibbard, and Kent S. Shelby. 2021. “Host Resistance to Bacillus Thuringiensis Is Linked to Altered Bacterial Community within a Specialist Insect Herbivore.” Molecular Ecology 30 (21): 5438–53. https://doi.org/10.1111/mec.15875. Palma, Leopoldo, Delia Muñoz, Colin Berry, Jesús Murillo, Primitivo Caballero, and Primitivo Caballero. 2014. “Bacillus Thuringiensis Toxins: An Overview of Their Biocidal Activity.” Toxins 6 (12): 3296–3325. https://doi.org/10.3390/toxins6123296. Paradis, Emmanuel, and Klaus Schliep. 2019. “Ape 5.0: An Environment for Modern Phylogenetics and Evolutionary Analyses in R.” Bioinformatics 35 (3): 526–28. https://doi.org/10.1093/bioinformatics/bty633. R Core Team. 2024. “R: A Language and Environment for Statistical Computing.” Vienna, Austria: R Foundation for Statistical Computing. https://www.R-project.org/. Raymond, Ben, Paul R. Johnston, Christina Nielsen-LeRoux, Didier Lereclus, and Neil Crickmore. 2010. “Bacillus Thuringiensis: An Impotent Pathogen?” Trends in Microbiology 18 (5): 189–94. https://doi.org/10.1016/j.tim.2010.02.006. Ripley, Brian, and Bill Venables. 2009. “MASS: Support Functions and Datasets for Venables and Ripley’s MASS.” https://doi.org/10.32614/CRAN.package.MASS. Roth, Louis M., and Ruth B. Howland. 1941. “Studies on the Gaseous Secretion of Tribolium Confusum Duval I. Abnormalities Produced in Tribolium Confusum Duval by Exposure to a Secretion given off by the Adults.” Annals of the Entomological Society of America 34 (1): 151–75. https://doi.org/10.1093/aesa/34.1.151. RStudio Team. 2023. “RStudio: Integrated Development for R.” PBC, Boston, MA URL: RStudio. www.rstudio.com. Sawada, Mitsuki, Takuma Sano, Kento Hanakawa, Patchara Sirasoonthorn, Takao Oi, and Ken Miura. 2020. “Benzoquinone Synthesis-Related Genes of Tribolium Castaneum Confer the Robust Antifungal Host Defense to the Adult Beetles through the Inhibition of Conidial Germination on the Body Surface.” Journal of Invertebrate Pathology 169 (January):107298. https://doi.org/10.1016/j.jip.2019.107298. Schwenke, Robin A., Brian P. Lazzaro, and Mariana F. Wolfner. 2016. “Reproduction-Immunity Trade-Offs in Insects.” Annual Review of Entomology 61:239–56. https://doi.org/10.1146/annurev-ento-010715-023924. Therneau, Terry M., Thomas Lumley (original S.->R port and R. maintainer until 2009), Atkinson Elizabeth, and Crowson Cynthia. 2024. “Survival: Survival Analysis.” https://cran.r-project.org/web/packages/survival/index.html. Therneau, Terry M., Patricia M. Grambsch, and V. Shane Pankratz. 2003. “Penalized Survival Models and Frailty.” Journal of Computational and Graphical Statistics 12 (1): 156–75. https://doi.org/10.1198/1061860031365. Vera-Ponce de León, Arturo, Benjamin C. Jahnes, Alejandro Otero-Bravo, and Zakee L. Sabree. 2021. “Microbiota Perturbation or Elimination Can Inhibit Normal Development and Elicit a Starvation-Like Response in an Omnivorous Model Invertebrate.” mSystems 6 (4): e0080221. https://doi.org/10.1128/mSystems.00802-21. Wang, Zhengyan, Haisheng Hu, Zhenzhen Chang, Shan Zhang, and Yujie Lu. 2023. “Molecular Mechanism of Enterococcus Faecalis-Induced Phosphine Sensitivity in Tribolium Castaneum (Coleoptera: Tenebrionidae).” Journal of Economic Entomology 116 (5): 1820–29. https://doi.org/10.1093/jee/toad171. Wickham, Hadley, Winston Chang, Lionel Henry, Thomas Lin Pedersen, Kohske Takahashi, Claus Wilke, Kara Woo, et al. 2024. “Ggplot2: Create Elegant Data Visualisations Using the Grammar of Graphics.” https://cran.r-project.org/web/packages/ggplot2/index.html. Yew, Joanne Y., and Henry Chung. 2015. “Insect Pheromones: An Overview of Function, Form, and Discovery.” Progress in Lipid Research 59 (July):88–105. https://doi.org/10.1016/j.plipres.2015.06.001. Yezerski, Ann, Carla Ciccone, Jill Rozitski, and Beth Volingavage. 2007. “The Effects of a Naturally Produced Benzoquinone on Microbes Common to Flour.” Journal of Chemical Ecology 33 (6): 1217–25. https://doi.org/10.1007/s10886-007-9293-2. Zhang, Husen, Joshua B. Sparks, Saikumar V. Karyala, Robert Settlage, and Xin M. Luo. 2015. “Host Adaptive Immunity Alters Gut Microbiota.” ISME Journal 9 (3): 770–81. https://doi.org/10.1038/ismej.2014.165. Additional Declarations The authors declare no competing interests. Supplementary Files SchulzandKorsaetalpreprintSI.pdf Supplementary Information Cite Share Download PDF Status: Posted 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-6353988","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":436943123,"identity":"312a8029-46d9-4567-94c9-a82b5f0e9346","order_by":0,"name":"Nora K.E. Schulz","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABD0lEQVRIie3QsUrEMBjA8a8EkiXl1pR7BiFwUO/g6LOkBOpS7HCj4N2UW9RZB9/BSRwLgXbpAxRuuS7ZhE5SoaAtDiqk6uiQ/5aQH18SAJfrHyeAgHcEYN83fyYIEP8k4m8Es6/rSXJKrpsW+uica2Iu/KclnNzKonnt1hkQmdvI6qpcME/JDdeQHPyKQVgnZwsqktWOGusYXifAvF0eP2goDr5i28c6DecgNAeW8gmCOuhH4qnNQIYp2UvQibeBZO0EGV6NR4Iw+iApZlTk4xTr83lV4GWsZHynMQruR1KZcE4TyTE19ouVCtVtH8U35b5pn9UlhKU0QbeO+IzIo3XMmPVjAE+ed7lcLtdvvQM6B1sl3HObWgAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0000-0003-3681-720X","institution":"University of Münster","correspondingAuthor":true,"prefix":"","firstName":"Nora","middleName":"K.E.","lastName":"Schulz","suffix":""},{"id":436943124,"identity":"d879ede1-1ec7-4566-be85-ddf7f1d7f057","order_by":1,"name":"Ana Korsa","email":"","orcid":"https://orcid.org/0000-0002-8050-0464","institution":"University of Münster","correspondingAuthor":false,"prefix":"","firstName":"Ana","middleName":"","lastName":"Korsa","suffix":""},{"id":436950658,"identity":"0c33308c-c826-4a5e-a4f9-0a06240d1f84","order_by":2,"name":"Helle Jensen","email":"","orcid":"","institution":"University of Münster","correspondingAuthor":false,"prefix":"","firstName":"Helle","middleName":"","lastName":"Jensen","suffix":""},{"id":436950659,"identity":"5091b3e6-5498-46e7-92dc-9049bfa75adb","order_by":3,"name":"Lai Ka Lo","email":"","orcid":"https://orcid.org/0000-0002-8289-9229","institution":"Max Planck Institute for Chemical Ecology","correspondingAuthor":false,"prefix":"","firstName":"Lai","middleName":"Ka","lastName":"Lo","suffix":""},{"id":436950660,"identity":"ae761835-fa6e-4225-8dfc-ef818fa9daf3","order_by":4,"name":"Jeanne Friedrichs","email":"","orcid":"","institution":"Bielefeld University","correspondingAuthor":false,"prefix":"","firstName":"Jeanne","middleName":"","lastName":"Friedrichs","suffix":""},{"id":436950661,"identity":"13e24d23-fac7-4867-8516-0d7af586b468","order_by":5,"name":"Caroline Müller","email":"","orcid":"https://orcid.org/0000-0002-8447-534X","institution":"Bielefeld University","correspondingAuthor":false,"prefix":"","firstName":"Caroline","middleName":"","lastName":"Müller","suffix":""},{"id":436950662,"identity":"4d533a82-57e4-4591-aabf-e63a4748c717","order_by":6,"name":"Joachim Kurtz","email":"data:image/png;base64,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","orcid":"https://orcid.org/0000-0002-7258-459X","institution":"University of Münster","correspondingAuthor":true,"prefix":"","firstName":"Joachim","middleName":"","lastName":"Kurtz","suffix":""}],"badges":[],"createdAt":"2025-04-01 14:15:01","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-6353988/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6353988/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":79730390,"identity":"abb797bc-3a9a-4547-a4aa-8f326a2f1ae3","added_by":"auto","created_at":"2025-04-02 05:33:46","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":150020,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePhenotypic readouts. \u003c/strong\u003e\u0026nbsp;\u003cstrong\u003eA)\u003c/strong\u003e Hazard ratios of mortality after \u003cem\u003eBtt\u003c/em\u003e exposure of different \u003cem\u003eT. castaneum\u003c/em\u003e experimental evolution regimes compared to the mortality of the \u003cem\u003eniche/pathogen-free\u003c/em\u003eregime (showing the median of the four replicate lines per selection regime with 95% confidence intervals, n=96 per line). A hazard ratio smaller than 1 means lower mortality probability during the four days observed than the control group. \u003cstrong\u003eB)\u003c/strong\u003e Developmental progression ratio showing the probability of pupation between 19 to 32 days post-oviposition relative to the control regime (showing the median of the four replicate lines per regime with 95% confidence intervals, n=48 per line). Ratios higher than 1 indicate a faster development to the pupal stage compared to the control. \u003cstrong\u003eC) \u003c/strong\u003eEarly life fecundity relative to the control regime (showing the mean number of offspring per pair, averaged across all four replicate lines with 95% confidence intervals, n = 20 pairs per replicate line). \u003cstrong\u003eD)\u003c/strong\u003e Biplot showing the results of a principal component analysis (PCA; variance explained by PC1 and 2) based on host life history traits, “Survival” (survival probability upon bacterial exposure), “Developmental time” (day to pupation) and “Fecundity” (number of live larvae produced) for generation 18, across selection lines, with each dot representing one of the four replicate lines. Data for generation 3-9 (shaded in grey) taken from Lo and Schulz et al. (2025).\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-6353988/v1/7cc7d20e733cb6c20304c8c6.png"},{"id":79730387,"identity":"0dffe8eb-43d2-400e-b77b-81e3d978b3a7","added_by":"auto","created_at":"2025-04-02 05:33:46","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":326775,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMicrobiota composition across regimes.\u003c/strong\u003e \u003cstrong\u003eA\u003c/strong\u003eBubble plot showing the relative abundance of the top 15 bacterial genera across experimental regimes and generations. Each point represents a genus, with bubble size indicating mean relative abundance. \u003cstrong\u003eB \u003c/strong\u003eChao1 richness estimates across regimes. Boxplots display the median and interquartile range. Whiskers extend from the box to the smallest and largest values within 1.5 times the IQR from the lower and upper quartiles, respectively. Jittered points represent individual samples of alpha diversity, with jittered points representing individual lines and their replication. Symbol represent generation 12 and 15. No significant differences were detected between regimes (Kruskal-Wallis, p \u0026gt; 0.05). \u003cstrong\u003eC\u003c/strong\u003e Principal Component Analysis (PCA) of Aitchison distance based on CLR-transformed OTU data. Points are coloured by regime and shaped by generation; ellipses represent 95% confidence intervals. Regimes show partial overlap, indicating weak separation of microbial community composition. \u003cstrong\u003eD\u003c/strong\u003eRandom Forest feature importance plot showing the top 20 OTUs contributing to classification of regimes based on microbiota composition. Bars represent mean decrease in classification accuracy, with taxa annotated at the genus level and coloured by regime association.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-6353988/v1/061aa8e9ecf46ca6675b6059.png"},{"id":79730379,"identity":"55cedd27-f15c-48f6-9333-0aeb3c8824e2","added_by":"auto","created_at":"2025-04-02 05:33:45","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":227443,"visible":true,"origin":"","legend":"\u003cp\u003eLinear discriminate analyses of stink gland secretion profiles of adult beetles (F2 after 18 generations of selection). \u003cstrong\u003eA \u003c/strong\u003eLDA of the four selection regimes in female beetles (n= 20 per selection regime)\u003cstrong\u003e B \u003c/strong\u003eLDA of the four selection regimes in male beetles (n= 20 per selection regime)\u003cstrong\u003e C \u003c/strong\u003eLDA grouping samples by “niche” selection treatment combined for both sexes (n=60)\u003cstrong\u003e D \u003c/strong\u003eC LDA grouping samples by “pathogen” selection treatment combined for both sexes (n=60)\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-6353988/v1/8ce6a6ded71471291549a6e3.png"},{"id":79730859,"identity":"14f8480a-7b11-478c-b395-4ed73787d6f8","added_by":"auto","created_at":"2025-04-02 05:41:51","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1427947,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6353988/v1/22bb1b4f-4e9f-49f7-a017-131d96938a1f.pdf"},{"id":79730384,"identity":"00e05d16-e970-4b7b-a94b-af9a82c2dc72","added_by":"auto","created_at":"2025-04-02 05:33:46","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":669398,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Information\u003c/p\u003e","description":"","filename":"SchulzandKorsaetalpreprintSI.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6353988/v1/1819946f0cc689909c5fdd83.pdf"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003e\u003cstrong\u003eResistance evolution driven by niche construction – the role of microbiota and beetle secretions\u003c/strong\u003e\u003c/p\u003e","fulltext":[{"header":"Introduction","content":"\u003cp\u003eMany organisms reshape their environments in ways that influence both their ecological interactions and evolutionary trajectories. This process, known as niche construction, can generate feedback loops that are suggested to impact an organism\u0026rsquo;s adaptation and survival (Laland, Odling-Smee, and Feldman \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e1999\u003c/span\u003e; Lo and Schulz et al. 2025). Such feedback is created by the inheritance of the changed niche from parents to offspring (called \u0026lsquo;ecological inheritance\u0026rsquo;), which, combined with genetic inheritance, could speed up the adaptation processes.\u003c/p\u003e \u003cp\u003eIn dense populations, where close contact increases the risk of disease transmission, these environmental modifications may play an important role in mediating or suppressing pathogen infections. For instance, many insects release different volatile and non-volatile chemical compounds to navigate social interactions and defend themselves against external threats. These secretions have different purposes: they can mediate communication and act as a defence against predators and parasites (Yew and Chung 2015). Many tenebrionid beetles have evolved exocrine glands that influence their external immune defences by producing a mixture containing different quinones ( Roth and Howland \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e1941\u003c/span\u003e; Joop et al. \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). Besides protection in the form of \u0026ldquo;external immunity\u0026rdquo; (Cotter and Kilner \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Otti, Tragust, and Feldhaar 2014), this emission of defensive secretions into the environment can also represent a niche construction (Odling-Smee, Laland, and Feldman 2013; M\u0026uuml;ller et al. 2020;). For example, defensive secretions from the stink glands of some beetle species (Cotter and Kilner \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2010\u003c/span\u003e) exhibit antimicrobial properties, thereby modifying the microbial environment in which these beetles reside (Yezerski et al. \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2007\u003c/span\u003e; Li et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). Regarding niche construction, both the defensive secretions in the environment and the microbial community of the environment or the organism itself can be transmitted from parent to offspring (K\u0026ouml;rner, Steiger, and Shukla \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Thus, they could represent potentially relevant components that contribute to ecological inheritance.\u003c/p\u003e \u003cp\u003eInsects can obtain their microbiota through both environmental sources as well as from parent to offspring (Geib et al. \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2009\u003c/span\u003e). A common form of environmental transfer in insects occurs through the consumption of feces from members of the same species, known as coprophagy. Studies on the red flour beetle (\u003cem\u003eTribolium castaneum\u003c/em\u003e, Tenebrionidae) have revealed that its microbiota reflects the flour microbiota shared across different life stages, with variations due to beetle population density and grain type, indicating the influence of coprophagy and diet (Agarwal and Agashe 2020; Bi, Zhang, and He 2024). Similar patterns have been observed in other insect species, where microbial transfer via coprophagy enhances host fitness (Jahnes, Herrmann, and Sabree \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Vera-Ponce de Le\u0026oacute;n et al. 2021), suggesting that such behaviours may serve an adaptive function. More broadly, acquisition of commensal microbes, whether through vertical or horizontal transmission, can offer significant benefits to the host, including improved immunity and fitness. For instance, despite variation in the microbiota of \u003cem\u003eT. castaneum\u003c/em\u003e across life stages and environmental conditions, it provides immune and fitness advantages and is further shaped by the host\u0026rsquo;s immune experiences (Futo, Armitage, and Kurtz 2016; Agarwal and Agashe 2020; Korša et al. \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). These observations underscore the significance of microbiome-mediated effects on ecological niches and host immunity, suggesting a complex interplay between genetic and ecological inheritance.\u003c/p\u003e \u003cp\u003eThe microbiome has also been shown to play a vital role in determining resistance (Zhang et al. \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2015\u003c/span\u003e), with certain gut bacteria influencing the host response to environmental stress and chemical threats. For example, specific bacteria like \u003cem\u003eEnterococcus faecalis\u003c/em\u003e can alter the physiological stress responses of \u003cem\u003eT. castaneum\u003c/em\u003e, thereby modulating its susceptibility to fumigants (Wang et al. \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). The speed at which resistance develops is greatly influenced by environmental factors, such as climate and the diversity of available habitats (Maino, Umina, and Hoffmann \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Moreover, the consumption of pathogen-infested food has been shown to alter quinone production levels, affecting the beetle's chemical defences against pathogens (Davyt-Colo et al. 2022).\u003c/p\u003e \u003cp\u003eA recent study provided insights in how niche construction affects selection, especially in the context of host-parasite interactions (Lo and Schulz et al. 2025). In this study, the effect of niche construction on resistance evolution and host adaptation to a pathogen was assessed using the experimental evolution of a well-established host-parasite model with \u003cem\u003eT. castaneum\u003c/em\u003e and the entomopathogenic bacterium, \u003cem\u003eBacillus thuringiensis tenebrionis\u003c/em\u003e (\u003cem\u003eBtt\u003c/em\u003e) (Lo and Schulz et al. 2025). Adult red flour beetles modify or construct their flour environment using potent antibacterial secretions from their stink glands (SGS), thereby providing external immune protection for both the adults and their offspring (Cotter et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Joop et al. \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; M\u0026uuml;ller et al. 2020). In this experimental evolution setup, experimental removal of niche construction\u0026mdash;using RNAi of a key gene needed to produce stink gland secretions\u0026mdash;was used to test the effects of the beetles\u0026rsquo; adaptation to \u003cem\u003eBtt\u003c/em\u003e. In each generation, beetles inhabited conditioned flour either with regular or drastically reduced stink gland secretion content. Importantly, as the RNAi treatment was performed in siblings of the beetles exposed to the niches, also the niche-constructing trait itself\u0026ndash;the SGS\u0026ndash;was allowed to evolve. After only three generations, populations with niche construction and exposure to the pathogen showed improved survival against pathogen infection compared to populations without niche construction (Lo and Schulz et al. 2025). By the ninth generation, this enhanced survival was apparent in all \u003cem\u003eBtt-\u003c/em\u003eselected populations but was particularly strong in those inhabiting constructed niches. Moreover, RNAseq of evolved beetles showed that gene expression differed strongly between the selection regimes, revealing that the mechanisms underlying resistance were dependent on niche construction. It is yet unclear whether the observed differences after nine generations of evolution are stable and will prevail for a longer period of selection, and what causes the observed differences caused by the experimental removal of niche construction. In addition to genetic adaptations achieving resistance\u0026ndash;and which might differ in the presence or absence of niche construction\u0026ndash;ecological inheritance could have driven the adaptation process. It is therefore, important to gain insights into two potentially relevant components of ecological inheritance: the SGS and the microbiota.\u003c/p\u003e \u003cp\u003eIn the present study, we continued this experimental evolution experiment and assessed the survival, development, and fitness measures of generations 12, 15, and 18. Furthermore, with 16S rRNA sequencing, we assessed the microbiota composition of the larvae in generations 12 and 15. We aimed to examine whether niche and pathogen selection shape overall stink gland secretion profiles. We thus examined the chemical secretions of the beetles from the different evolution regimes in the 18th generation using gas chromatography-flame ionization detection. Overall, this study sheds light on how organisms actively modify their environment and how these changes, in turn, impact their microbiome and chemical ecology.\u003c/p\u003e"},{"header":"Materials and methods","content":"\u003cp\u003e\u003cstrong\u003eNiche construction and experimental evolution design\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eTribolium castaneum\u003c/em\u003e beetles used in this study originated from experimentally evolved lines described in Lo and Schulz et al. (2025). Beetles were maintained in glass jars containing heat-sterilized (75 \u0026deg;C) organic wheat flour (type 550, Drogerie-Markt (DM) supplemented with 5% brewer\u0026rsquo;s yeast (hereafter referred to as \u0026ldquo;flour\u0026rdquo;). All lines were kept at 30 \u0026deg;C and 70% relative humidity under a 12-hour light/dark cycle.\u0026nbsp;In each generation, we used siblings from the previous generation as flour-conditioning beetles (\u003cem\u003eniche constructors\u003c/em\u003e) as described in Lo and Schulz et al. (2025).\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003eTo produce secretion-free lines, we impaired the proper development of stink glands and the subsequent production of SGS (\u003cem\u003eniche-free\u003c/em\u003e regimes) through pupal knockdown of\u0026nbsp;\u003cem\u003eDrak\u003c/em\u003e gene using RNAi (\u003cem\u003eDrak\u003c/em\u003e-dsRNA 1000 ng/\u0026micro;l in phosphate-buffered saline (PBS), Lo and Schulz et al 2025). For simplicity, we call this treatment \u0026lsquo;\u003cem\u003eniche-free\u0026rsquo;\u003c/em\u003e. For niche regimes, control beetles received\u0026nbsp;\u003cem\u003eeGFP-\u003c/em\u003edsRNA injections at the same concentration, which does not affect target gene expression.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe experimental evolution experiment consisted of four regimes as described in Lo and Schulz et al. (2025; Figure S1): two regime groups, each with four replicate lines, were subjected to pathogen selection after developing in either constructed or non-constructed niches (\u003cem\u003eniche/pathogen\u003c/em\u003e and\u0026nbsp;\u003cem\u003eniche-free/pathogen,\u003c/em\u003e respectively), and two regime groups, each with four replicate lines, developed in constructed or non-constructed niches without pathogen selection (\u003cem\u003eniche/pathogen-free\u003c/em\u003e and\u0026nbsp;\u003cem\u003eniche-free/pathogen-free\u003c/em\u003e, respectively). To start the selection experiment, freshly laid eggs were placed in jars containing flour that had been conditioned with 100 flour-conditioning beetles for four days to establish the initial generation. On the 15th day post-oviposition, 192 larvae per line were randomly chosen and exposed to either the pathogen infected or the control treatment. Following a four-day exposure period, 100 surviving larvae were selected and transferred to jars containing freshly conditioned flour (0.2 g per beetle). This process of exchanging conditioned flour continued until the selection line individuals emerged as adults and began producing their secretions. After one to two months, these adult beetles were used to generate the subsequent generation of\u0026nbsp;\u003cem\u003eniche constructors\u003c/em\u003e and inhabitants.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePathogen infection assays\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFor the selection and survival assays,\u0026nbsp;\u003cem\u003eBacillus thuringiensis tenebrionis\u003c/em\u003e (\u003cem\u003eBtt\u003c/em\u003e; BGSCID 4AA1) was obtained from the Bacillus Genetic Stock Centre (BGSC, Ohio State University, USA) and the assays were prepared as described in Lo and Schulz et al. (2025). Methods for cultivation and oral infection were adapted from Milutinović et al. (2013). To prepare overnight cultures,\u0026nbsp;\u003cem\u003eBtt\u003c/em\u003e from glycerol stock was introduced into 3-4 ml of Bt culture medium, including 15 \u0026mu;l of filter-sterilized salt solution (using 0.2 \u0026mu;m cellulose nitrate filters, Whatman) and 3.75 \u0026mu;l of 1M CaCl\u003csub\u003e2\u003c/sub\u003e. This was incubated at 30 \u0026deg;C and 180 rpm. The next day, 400 mL of Bt medium, supplemented with 1 mL of salt solution and 500 \u0026mu;l of CaCl\u003csub\u003e2\u003c/sub\u003e, was inoculated with the overnight culture and incubated in darkness at 30 \u0026deg;C and 180 rpm for eight days until sporulation was complete.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAfter an eight-day sporulation period, the spores were rinsed and diluted to 2 x 10^10 spores/ml using phosphate-buffered saline (PBS, Calbiochem). This suspension was then combined with a heat-sterilized flour mixture at a ratio of 0.15 g per mL of spore suspension. Under sterile conditions, 10 \u0026mu;L of this mixture was dispensed into each well of flat-bottom 96-well plates (Sarstedt), sealed with a breathable adhesive film (Kisker Biotech GmbH \u0026amp; Co. KG) and left to dry overnight at 30\u0026deg;C in darkness. For the regime groups without\u0026nbsp;\u003cem\u003eBtt\u003c/em\u003e exposure (\u003cem\u003eniche/pathogen-free\u003c/em\u003e and\u0026nbsp;\u003cem\u003eniche-free/pathogen-free\u003c/em\u003e), the plates were prepared identically, but without bacteria.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePhenotypic readouts in generations 12, 15 and 18\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDuring generations 12, 15, and 18 of continuous experimental evolution focusing on adaptation to pathogens in both\u0026nbsp;\u003cem\u003eniche\u003c/em\u003e and\u0026nbsp;\u003cem\u003eniche-free\u003c/em\u003e environments, we conducted phenotypic evaluations across all replicate lines following two generations of relaxed selection, referred to as the first (F1) and second (F2) relaxation generations as detailed in Lo and Schulz et al. (2025). We evaluated survival rates post-pathogen infection, the speed of development until pupation, and early-life reproductive capacity. All experimental subjects were derived from two sets of 24-hour egg-laying periods by 100 F1 (from generations 12,15, and 18) adults in glass containers, following previously established protocols.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSurvival after pathogen infection\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo assess the development of resistance to\u0026nbsp;\u003cem\u003eBtt\u003c/em\u003e, we subjected 96 larvae per line from all lines across the four selection regimes to a concentration of 2 x 10\u003csup\u003e10\u003c/sup\u003e spores/ml of\u0026nbsp;\u003cem\u003eBtt\u0026nbsp;\u003c/em\u003e(LD50 in the original population) and monitored their survival over four days (for detailed methods see Lo and Schulz et al., 2025). The oral infection of\u0026nbsp;\u003cem\u003eBtt\u003c/em\u003e for the survival test followed the same methodology as the selection protocol. We randomly selected F2 larvae from each selection line at 15 days of age, with 96 individuals per replicate line exposed to\u0026nbsp;\u003cem\u003eBtt\u003c/em\u003e spores. Additionally, we randomized larvae from different regimes across different plates to avoid potential plate effects.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDevelopmental assay\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo assess the impact of niche construction on development time, we monitored the pupation timing of F2 beetles raised under unrestricted feeding conditions without exposure to pathogens (for detailed methods see Lo and Schulz et al., 2025). At 14 days after egg-laying (dpo), we placed 48 larvae per selection line individually into 96-well plates. Each well contained roughly 0.2 g of flour diet. The larvae were maintained under standard rearing conditions for the duration of the experiment. We recorded pupation daily from 19 dpo through 32 dpo.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFecundity\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo test for potential costs of evolved resistance or adaptation to the constructed niches, we measured fecundity of F2 beetles as a proxy of reproductive fitness (for detailed methods see Lo and Schulz et al., 2025). 24 days after oviposition, the majority of beetles from a single F1 beetle oviposition had reached the pupal stage. We identified their sexes and placed the F2 pupae individually. After all beetles had reached sexual maturity, we placed 20 pairs of one male and one female from each line into vials with 5 g of flour for three days. Following this interval allowing for mating, the adult beetles were removed from the flour. Twelve days later, we counted the number of living larvae in each vial.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStatistical analysis of phenotypic readouts\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll analysis were performed in R (R Core Team 2024; version 4.4.2) using R Studio (RStudio Team 2023; 2023; Posit Software, version 2024.12.0). For survival and development, data we performed Cox proportional hazard analysis using the \u0026ldquo;survival\u0026rdquo; package (version 2.38; Therneau et al. 2024; Therneau, Grambsch, and Pankratz 2003) on each generation separately. Hazard ratios, or progression ratios, respectively, were calculated compared to the internal control, the \u003cem\u003eniche/pathogen-free\u003c/em\u003e regime, while pooling replicate lines for the analysis. For the fecundity data analysis, we excluded pairs with less than five offspring, as we assume that these occurred rather due to technical reasons (premature death of one parent from handling, misassignment of sex, etc.) than that they were of biological relevance. A GLMM with negative binomial error distribution was fitted to the remaining data using the \u0026ldquo;lme4\u0026rdquo; package (Bates et al. 2015).\u0026nbsp;The number of larvae served as response variable, replicate line as random factor and\u0026nbsp;selection regime as explanatory variable with the three regimes compared against the internal control \u003cem\u003eniche/pathogen-free\u003c/em\u003e regime.\u0026nbsp;On generation 18, we additionally performed a principal component analysis (PCA) to visualize how the three phenotypic readouts (survival of infection, pupation rate, and fecundity) contribute to the differences between regimes and how they relate to each other. Principal components were calculated with the standardized means for each of the four replicate lines per selection regime using the \u0026ldquo;prcomp\u0026rdquo; function and results were visualized using \u0026ldquo;ggbiplot\u0026rdquo; function from the ggplot2 package\u0026nbsp;(Wickham et al. 2024).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMicrobial community profiles\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo examine the impact of selection regimes on the microbiota of the beetles, we analysed the F2 of evolved beetle lines from generations 12 and 15. We took three biological replicates from each selection line within both generations, resulting in 96 samples. RNA was extracted from pools of larvae (see below) and transcribed to cDNA before performing 16S rRNA amplicon sequencing of the V3-V4 hypervariable region. RNA from biological replicates was extracted on different days to ensure randomization. We then clustered sequences to operational taxonomic units (OTUs).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRNA extraction and cDNA synthesis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFor each replicate, we pooled twenty 14-day-old larvae, transferred them to sterilized 1.5 ml Eppendorf tubes, froze them in liquid nitrogen and stored them at -80 \u0026deg;C. Generation 12 larvae were stored for approximately one year, while generation 15 larvae were extracted within a week. Using a sterile pestle, larvae were ground in liquid nitrogen. Half of the sample was transferred to a new tube with 100 \u0026mu;l phenol-chloroform-isoamyl alcohol and briefly mixed. Subsequent steps followed the RNeasy Power Microbiome Kit (QIAGEN N.V., Netherlands) protocol, with minor alterations. We added 70% ethanol to avoid co-purification of small RNA species. RNA was eluted in 50 \u0026mu;l of nuclease-free water. RNA concentration and quality were measured with a NanoPhotometer (Pearl, Implen, USA), Qubit BR RNA assay (Invitrogen Qubit Fluorometer 2.0, Thermo Fisher Scientific Inc., USA) and a subset with the Agilent 2100 Bioanalyzer for RNA integrity (Agilent Technologies Inc, USA). RNA was stored at -80 \u0026deg;C. For cDNA synthesis, the RevertAid II Kit (Thermo Fisher Scientific Inc., USA) with random hexamer primers was used, adding 5 \u0026mu;l RNA per reaction. Samples were transported to the sequencing facility on dry ice.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e16S rRNA sequencing and processing\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe library preparation and sequencing of 16S rRNA was executed by Novogene (Novogene UK, Cambridge Sequencing Center). Amplification of cDNA was performed with 10 ng template DNA following PCR purification using magnetic beads. The V3-V4 hypervariable region was amplified (primers: CCTAYGGGRBGCASCAG; GGACTACNNGGGTATCTAAT) and sequenced on a paired-end Illumina platform to generate 250bp paired-end raw reads. Splitting of paired-end reads and truncation of barcodes and primers was performed using Python (V3.6.13) and cutadapt (V3.3). Using FLASH (V1.2.11) and fastp (V0.23.1) the paired-end reads were merged and filtered to obtain high-quality merged sequences. Chimeric sequences were removed with vsearch (V2.16.0).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eOperational taxonomic units (OTUs) were clustered using Uparse software (V7.01001), with sequences \u0026ge; 97% similar assigned to the same OTUs. Species annotation used the Mothur algorithm applied to the SILVA 138.1 database (Quast et al. 2013). The SILVA reference was filtered to cluster 99% similar sequences for accurate taxonomy assignments. As SILVA only provides species-level taxonomy, NCBI searches supplemented information on other taxonomic levels. MUSCLE software (V3.8.31) constructed phylogenetic relationships through multiple sequence alignment. Bioperl\u0026apos;s (V1.7) SVG function constructed the phylogenetic tree.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAnalysis of microbiota data\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eR (V4.3.3, R Core Team 2024) and R studio (V 2023.06.0 + 421, RStudio Team 2023) were used to conduct all data visualization and analysis. Taxonomy, sample information and OTU count data were merged using the package \u0026ldquo;phyloseq\u0026rdquo; in R (McMurdie and Holmes 2013) with the help of the packages \u0026ldquo;biomformat\u0026rdquo; (McMurdie \u0026amp; Paulson 2017), \u0026ldquo;ape\u0026rdquo; (Paradis and Schliep 2019) and \u0026ldquo;Matrix\u0026ldquo; (Bates, Maechler, and Jagan 2000). Low abundant taxa were filtered out by removing OTUs that occurred with less than ten copies in individual samples, as described in (Nikodemova et al. 2023). In the next step, Cyanobacteria, Archaea and unassigned OTUs were removed using the functions available in \u0026ldquo;phyloseq\u0026rdquo;.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAlpha diversity was assessed using the Chao1 richness estimator to evaluate differences in microbial richness across experimental regimes. Chao1 values were calculated using the \u0026ldquo;estimate_richness()\u0026rdquo; function from the \u0026ldquo;phyloseq\u0026rdquo; package. To compare richness across regimes, a Kruskal-Wallis rank sum test was applied, as the data did not meet the assumptions of parametric testing.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eBeta diversity was assessed using Aitchison distance (Aitchison et al. 2000), which accounts for the compositional nature of microbiome data, using the \u0026ldquo;vegan\u0026rdquo; package (Oksanen et al. 2001). The OTU table was CLR-transformed using the \u0026ldquo;clr()\u0026rdquo; function from the \u0026ldquo;compositions\u0026rdquo; R package after applying a pseudocount of 1 to avoid issues with zero counts. PCA was performed on the clr-transformed data to visualize differences in microbial community composition across regimes and generations.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo test for differences in community composition between regimes, we used PERMANOVA (\u0026ldquo;adonis2\u0026rdquo; from the \u0026ldquo;vegan\u0026rdquo; package) with 999 permutations. Assumptions of homogeneity of group dispersions were assessed using \u0026ldquo;betadisper()\u0026rdquo; followed by an ANOVA. Pairwise comparisons between regimes were conducted using the \u0026ldquo;pairwise.adonis\u0026rdquo; function with Bonferroni-adjusted p-values. Furthermore, ANOSIM was performed to complement PERMANOVA results and evaluate the relative strength of group separation.\u003c/p\u003e\n\u003cp\u003eWe applied Random Forest classification to identify microbial taxa (OTUs) most important for distinguishing microbiota between experimental regimes. The OTU table was CLR-transformed using the \u0026ldquo;compositions\u0026rdquo; package after adding a pseudocount of 1 to account for zero inflation. Random Forest models were trained using the \u0026ldquo;randomForest\u0026rdquo; package (Breiman et al. 2002) in R with 1000 trees and default parameters. Feature importance was assessed using the Mean Decrease in Accuracy metric. The top 20 OTUs were visualized in a horizontal bar plot, with taxonomy assigned from the phyloseq object and colored by the regime with which they were most associated. Taxonomic labels were derived at the genus level when available.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStink gland secretion profiles\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eExtraction method and data compilation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo extract stink gland secretions, groups of six beetles of the same sex were transferred to pre-cooled 2 ml microreaction tubes and placed in ice water for 3 min to induce excretion of gland secretion as described before (n = 5 samples per treatment group and sex; Joop et al. 2014; Lo et al. 2023). Beetles inside the tubes were freeze-killed directly after secretion at -20\u0026deg;C. For the extraction, we added 360 \u0026micro;l ice-cold acetone (\u0026ge; 99.8% HPLC grade, Fisher Scientific, Loughborough, UK) containing 0.05 mg/ml \u003cem\u003en\u003c/em\u003e-octadecane (\u0026ge; 98.5%, Sigma-Aldrich, St. Louis, USA) as internal standard. Samples were shaken for 5 min at 4\u0026deg;C, centrifuged and supernatants were transferred to vials. Replicates of the different treatments were prepared fresh and either directly analysed or stored for one day at -20\u0026deg;C. In addition, at each extraction day, one blank without beetles was prepared (total n = 5). Beetle samples and blanks were analysed using a gas-chromatograph connected to a flame ionization detector (GC-2010 Plus-FID, Shimadzu, Kyoto, Japan) equipped with a VF-5ms column (30 m \u0026times; 0.25 mm \u0026times; 0.25 \u0026mu;m, 10 m guard column, Varian, Agilent Technologies, Santa Clara, California, USA), with nitrogen as carrier gas. The samples were injected at 250\u0026deg;C at a split of 5 with a constant column flow rate of 1 ml/min. The temperature program started at 50\u0026deg;C for 2 min, was then increased in 10\u0026deg;C/min steps to 300\u0026deg;C and the final temperature was maintained for 10 min. Alkane standard mixtures C7-40 and C21-C40 (both Sigma-Aldrich, St. Louis, USA) were analysed using the same temperature program to calculate retention indices (Kov\u0026aacute;ts 1958; van Den Dool and Dec. Kratz 1963). In addition, standards (30 ng/\u0026micro;l) of 2-methyl-1,4-benzoquinone (MBQ), 2-methyl-1,4-hydroquinone (MHQ), 1-pentadecene (all 98%, Sigma-Aldrich) and 2-ethyl-1,4-hydroquinone (EHQ, 97%, Abcr GmbH, Karlsruhe, Germany) were measured for metabolite verification.\u003c/p\u003e\n\u003cp\u003ePre-processing of the data such as peak integration was done in GCSolution Postrun analysis (Version 2.30, update 2.41, Shimadzu, Kyoto, Japan), with the following parameters: width of 3 sec., slope of 1200 uV/min, min. area/height of 0 counts and default band time of 0.03 min. Metabolites of interest were extracted automatically or integrated manually. Alignment of the retention times was performed in R (version 4.4.1, (R Core Team 2018)) and RStudio (version 2024.04.2+764, R Studio Team 2024) using the package \u0026ldquo;GCalignR\u0026rdquo; (Ottensmann et al. 2018) with a maximum linear retention time shift (max_linear_shift) of 0.03 min. Peak areas of the metabolites were normalized to the area of the internal standard. Only metabolites were further considered, whose areas were at least five times higher than the averaged areas in the blank samples.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo identify metabolites, eight samples of the different treatment groups and sexes, the two alkane series and the standards were analysed using a GC coupled to a mass spectrometer (GC 2010 Plus \u0026ndash; MS QP2020, Shimadzu, Kyoto, Japan) in electron impact ionization mode with an identical column as for the GC-FID and helium as carrier gas, but a split of 8. The temperature program was the same as used for the GC-FID measurements. The ion source temperature was set to 230\u0026deg;C and the interface temperature to 250\u0026deg;C. In quadrupole MS mode at 70 eV the line spectra were acquired with \u003cem\u003em\u003c/em\u003e/\u003cem\u003ez\u003c/em\u003e 40 to 600 and a scan event time of 0.25 sec. The mass spectra and retention indices were compared to those in the NIST database (NIST14, National Institute of Standards and Technology, Gaithersburg, Maryland) and Pherobas (El-Sayed 2024) for the putative identification. Metabolites measured by GC-FID were identified based on the RI.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData analysis of secretion profiles\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll data analysis and visualization were conducted using R (V4.3.3, R Core Team 2024) and R studio (V 2023.06.0 + 421, RStudio Team 2023). To test for differences in compound amounts\u0026mdash;represented by the normalized area of the GC peak\u0026mdash;we focused on five major representatives of stink gland secretions (SGS), the four quinones (EBQ, EHQ, MBQ, MHQ) and the carbohydrate 1-pentadecene. For each sex and metabolite separately, we fitted a linear mixed-effects model to the log-transformed data using the \u0026ldquo;lme4\u0026rdquo; package (Bates et al. 2015).\u003c/p\u003e\n\u003cp\u003eThe technical GC run was included as a random effect to account for variability across analytical replicates. Pairwise comparisons of model estimates were performed using the \u0026ldquo;emmeans\u0026rdquo; package (Lenth 2017).\u003c/p\u003e\n\u003cp\u003eTo explore potential differences in the multivariate composition of SGS profiles among selection regimes, we first performed non-metric multidimensional scaling (NMDS) using a Bray\u0026ndash;Curtis dissimilarity matrix, implemented in the \u0026ldquo;vegan\u0026rdquo; package (Oksanen et al. 2001). NMDS was conducted separately for each sex. This unsupervised ordination method provides a visual representation of sample relationships while making minimal distributional assumptions. As the NMDS plots did not reveal any group separation by regime, we followed up with a linear discriminant analysis (LDA) to formally assess whether group-level differences could be detected in a supervised classification framework. LDA was conducted using the \u0026ldquo;MASS\u0026rdquo; package (Ripley and Venables 2009) and scores for the first two linear discriminants (LD1 and LD2) were visualized using \u0026ldquo;ggplot2\u0026rdquo; (Wickham et al. 2024). Finally, we combined data across sexes and ran additional LDAs using only \u0026ldquo;niche\u0026rdquo; and \u0026ldquo;pathogen\u0026rdquo; as treatment factors, respectively. The resulting LD1 scores were visualized as density distributions to evaluate treatment-level separation.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003eEnhanced pathogen resistance continues to evolve without apparent costs\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe continued an evolution experiment testing the influence of niche construction (i.e. the presence or absence of defensive secretions) on pathogen adaptation and added another nine generations of experimental evolution (Figure S1, Lo and Schulz et. al 2025). Four replicate beetle lines had evolved with impaired \u003cem\u003evs.\u003c/em\u003e normal ability to construct their niches (for simplicity called \u0026lsquo;\u003cem\u003eniche-free\u0026rsquo;\u003c/em\u003e \u003cem\u003evs.\u003c/em\u003e \u0026lsquo;\u003cem\u003eniche\u0026rsquo;\u003c/em\u003e) and with \u003cem\u003evs.\u003c/em\u003e without \u003cem\u003eBtt\u003c/em\u003e exposure (called \u0026lsquo;\u003cem\u003epathogen\u0026rsquo;\u003c/em\u003e \u003cem\u003evs.\u003c/em\u003e \u0026lsquo;\u003cem\u003epathogen-free\u0026rsquo;\u003c/em\u003e). To test for continued evolution of resistance we measured survival of \u003cem\u003eBtt\u003c/em\u003e infection in the F2 offspring every third generation. The \u003cem\u003eniche/pathogen-free\u003c/em\u003e regime was considered the control, because it resembles the unselected situation, where beetles can construct their niche and do not encounter any pathogens. In the regimes with pathogen selection, host resistance continued to increase during the evolution experiment (Figure 1A; Table S1). After a total of 18 generations of selection, 92% (\u003cem\u003eniche/pathogen\u003c/em\u003e) and 94% (\u003cem\u003eniche-free/pathogen\u003c/em\u003e), respectively, survived the \u003cem\u003eBtt\u003c/em\u003e exposure in the pathogen selected regimes, while in the control (\u003cem\u003eniche/pathogen-free\u003c/em\u003e) the survival was only 57%. These survival differences demonstrate a significant decrease in hazard ratios, with the pathogen selection reducing mortality probabilities to 0.15 (\u003cem\u003eniche/pathogen\u003c/em\u003e; CI= 0.099-0.217; p\u0026lt;0.001; Figure 1A; Table S1) and 0.11 (\u003cem\u003eniche-free/pathogen\u003c/em\u003e; CI=0.072-0.172; p\u0026lt;0.001) compared to the \u003cem\u003eniche/pathogen-free\u003c/em\u003e regime. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAcross three time points (generations 12, 15, and 18), F2 offspring from the \u003cem\u003eniche-free/pathogen-free\u003c/em\u003e regime exhibited lower mortality than those from the \u003cem\u003eniche/pathogen-free\u003c/em\u003e control (Figure 1A; Table S1). Hazard ratios ranged from 0.87 to 0.72 (Table S1), with significantly lower values in generations 12 and 18 (p\u0026lt;0.05; Table S1). This indicates that resistance can emerge even without \u003cem\u003eBtt\u003c/em\u003e selection pressure when beetles are prevented from constructing their niche.\u003c/p\u003e\n\u003cp\u003eIn contrast to survival, we did not find any consistent changes in development over the course of the experiment (Figure 1B). In generation 12, all three regimes had a transiently faster development to pupation compared to the \u003cem\u003eniche/pathogen-free\u003c/em\u003e control, with the progression ratio (PR, i.e. probability to pupate over time relative to control) being more than twice as high in the pathogen-selected regimes (\u003cem\u003eniche/pathogen\u003c/em\u003e: PR=2.01, p\u0026lt;0.001; \u003cem\u003eniche-free\u003c/em\u003e/\u003cem\u003epathogen\u003c/em\u003e: PR=2.46, p\u0026lt;0.001). The larvae within the \u003cem\u003eniche-free/pathogen-free\u003c/em\u003e regime also pupated significantly earlier (PR: 1.57, p\u0026lt;0.001). However, this faster development disappeared in generation 15 with all four regimes pupating at a similar time (Figure 1B; Table S1). In generation 18, finally, only individuals of the \u003cem\u003eniche/pathogen\u003c/em\u003e (PR = 1.49; p\u0026lt;0.001) and \u003cem\u003eniche-free/pathogen-free\u003c/em\u003e regime (PR = 1.16; p\u0026lt;0.05) developed significantly faster than individuals of the \u003cem\u003eniche/pathogen-free\u003c/em\u003e control.\u003c/p\u003e\n\u003cp\u003eSimilarly, for fecundity we did not observe any consistent changes over the course of the evolution experiment (Figure 1C). Only in generation 18 did the pathogen-selected regimes show a decrease in reproduction, with 18% (\u003cem\u003eniche/pathogen\u003c/em\u003e) and 9% less offspring (\u003cem\u003eniche-free/pathogen\u003c/em\u003e; Table S1) hinting at possible costs of the evolved resistance.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFinally, we performed a principal component analysis (PCA) combining the three phenotypic readouts from generation 18 (Figure 1D). Principal Component 1 (PC1), accounting for over 50% of the variation, was primarily driven by mortality and fecundity differences and clearly separated the regimes based on pathogen selection. Principal Component 2 (PC2) explained 33% of the variation, mainly reflecting variation in development time, but did not align with any specific selection regime.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eThe microbial community differs between some regimes\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo determine whether \u003cem\u003eBtt\u003c/em\u003e selection and evolved resistance in combination with differentially conditioned niches affect the microbiome composition and diversity, we examined the V3-V4 region of bacterial 16S rRNA. We collected larvae from the F2 beetles of generation 12 and generation 15. For this, OTUs were clustered from raw Illumina Amplicon sequencing reads. After filtering, 17,794,626 reads remained with an average of 183,450 reads per sample across 95 samples. Of 6,875 OTUs in the original dataset, 4,487 OTUs remained after filtering. Of those remaining OTUs, only 11.03% could be assigned at the species level, 48.2 % at the genus level, and 74.4% at the family level. The most common genera were largely consistent between generations, with some fluctuations in their abundance (Figure 2A). However, since it has been previously shown that the microbiota of \u003cem\u003eT. castaneum\u003c/em\u003e is dependent on its environment (Agarwal and Agashe 2020), fluctuations between generations were expected. To focus on regime effects and potential selection-driven changes in the microbiota, we analysed the generations jointly.\u003c/p\u003e\n\u003cp\u003eMicrobial richness, as measured by the Chao1 index did not differ significantly across regimes (Figure 2B; Kruskal-Wallis \u0026chi;\u0026sup2; = 4.40, df = 3, p = 0.221). A PCA of Aitchison distance revealed weak clustering by regime, with some overlap between niche and pathogen treatments (Figure 2C). PERMANOVA indicated a statistically significant effect of regime on microbial community composition (R\u0026sup2; = 0.095, p = 0.001); however, a significant result from the PERMDISP test (F = 11.02, p \u0026lt; 0.001) suggested differences in within-group dispersion, meaning that although compositional differences exist between regimes, variation in group dispersion may partially contribute to the observed significance. Post hoc Tukey\u0026apos;s HSD tests revealed that \u003cem\u003eniche-free\u003c/em\u003e regimes had significantly lower dispersion and different microbial composition compared to \u003cem\u003eniche\u003c/em\u003e regimes, while pathogen-selected regimes did not differ in composition and dispersion within \u003cem\u003eniche\u0026nbsp;\u003c/em\u003eor \u003cem\u003eniche-free\u003c/em\u003e regime (Figure 2C; Table S2). ANOSIM also detected a significant but weak separation in microbiota between regimes (R = 0.10, p = 0.001), consistent with high variability in community structure.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eRandom Forest classification was used to identify the most important OTUs contributing to differences in microbial community composition across regimes (Figure 2D). The model was trained on CLR-transformed OTU data using 1000 trees and achieved an overall out-of-bag (OOB) error rate of 45.26%. Classification accuracy varied across regimes, with higher accuracy observed for microbiota of the \u003cem\u003eniche/pathogen\u003c/em\u003e and \u003cem\u003eniche/pathogen-free\u0026nbsp;\u003c/em\u003eregimes and lower performance for those of the \u003cem\u003eniche/pathogen-free\u0026nbsp;\u003c/em\u003eindividuals, which exhibited frequent misclassification. The top 20 OTUs were selected based on Mean Decrease in Accuracy and visualized with genus-level taxonomic annotations. These taxa, representing diverse bacterial groups, were identified as key contributors to the observed variation between regimes, highlighting specific microbial signatures associated with each selection treatment.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eThe selection regime had only minor effects on stink gland secretions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn the stink gland secretions collected from adult F2 beetles of generation 18, we detected 26 compounds and identified 22 of them, which produced a measurable signal in at least 25% of the samples, including the four quinones of interest\u0026mdash;ethylbenzoquinone (EBQ), methylbenzoquinone (MBQ), ethylhydroquinone (EHQ), and methylhydroquinone (MHQ)\u0026mdash;as well as the carrier compound 1-pentadecene (Table S3).\u003c/p\u003e\n\u003cp\u003eBecause SGS production is known to differ between sexes in \u003cem\u003eT. castaneum\u003c/em\u003e (Lo et al. 2023), we analysed the five focal compounds separately for females and males (Figure S2). Incorporating block (GC run) as a random effect due to considerable run-to-run variation, we found no significant differences in amounts of any of the five compounds between selection regimes for either sex (Figure S2, Table S4). This suggests that experimental evolution did not lead to measurable changes in these characteristic SGS compounds.\u003c/p\u003e\n\u003cp\u003eTo test whether selection regimes had instead driven changes in SGS composition, we applied NMDS using Bray\u0026ndash;Curtis dissimilarity matrices to the male and female datasets including all 26 compounds. In both cases, the 95% confidence ellipses around group centroids overlapped completely in the NMDS space (Figure S3), indicating no clear separation among selection regimes. As a follow-up, we conducted LDA. However, no consistent separation among selection regimes was observed in either the female (Figure 3A) or male (Figure 3B) LDA results. Finally, we combined the datasets across sexes and performed two additional LDAs using either \u0026ldquo;niche\u0026rdquo; (Figure 3C) or \u0026ldquo;pathogen\u0026rdquo; (Figure 3D) as the treatment factor. Again, no distinct group separation was detected along the first linear discriminant in either case.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eRecent findings demonstrated that niche construction can facilitate rapid adaptation to pathogen pressure (Lo and Schulz et al. 2025). Building on this, we continued to experimentally evolve beetle populations under different regimes\u0026mdash;either with or without the possibility to construct a niche, and with or without exposure to a pathogen\u0026mdash;to investigate whether evolution led to changes in resistance, larval microbiomes or adult secretion composition. Host resistance continued to evolve towards very high levels of resistance in regimes with pathogen exposure. We observed differences in microbiota profiles of larvae from the \u003cem\u003eniche\u003c/em\u003e versus \u003cem\u003eniche-free\u003c/em\u003e regimes, suggesting that niche construction fosters greater microbial variation. In contrast, chemical secretion profiles of adults showed no significant changes across regimes, indicating that these secretions remained stable throughout the selection process.\u003c/p\u003e \u003cp\u003eAdaptation to the pathogen increased both with and without the constructed niche during generations 12, 15, and 18 in \u003cem\u003eT. castaneum\u003c/em\u003e. This evolved resistance, which resulted in a greater survival after infection compared to \u003cem\u003eniche/pathogen-free\u003c/em\u003e regime, aligns with findings from earlier experimental evolution studies involving the same pathogen but other host species (Masri et al. \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Ferro et al. \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). The previous study of the same system (Lo and Schulz et al. 2025) used transcriptomic analyses after nine generations of experimental evolution to reveal that similar levels of resistance can arise through distinct mechanisms. Beetles that evolved pathogen resistance without a constructed niche exhibited constitutive changes in gene expression, even in the absence of pathogen exposure, suggesting a form of general, always-active resistance. In contrast, beetles that evolved in the presence of a constructed niche showed inducible expression changes, activated only upon \u003cem\u003eBtt\u003c/em\u003e exposure. The continued evolution in both regimes that we now observed suggests that both evolutionary trajectories can lead to further increase in resistance. While gaining resistance to pathogens often involves trade-offs in life history traits, as found in other insect species (Cotter and Kilner \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Schwenke, Lazzaro, and Wolfner 2016), both pathogen-selected regimes exhibited nearly complete resistance with seemingly minimal associated costs in the present study. However, it is important to note that in generation 18, the individuals of the \u003cem\u003eniche/pathogen\u003c/em\u003e regime generated approximately 9\u0026ndash;18% fewer offspring than those of the \u003cem\u003eniche-free/pathogen-free\u003c/em\u003e and \u003cem\u003eniche/pathogen free\u003c/em\u003e regimes (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eC). Under natural conditions, even if low, such a reproductive cost would maybe result in non-resistant lines outcompeting and eventually replacing resistant ones in the absence of continued pathogen pressure. Additionally, there could be some hidden costs that we did not capture in our readouts, such as larval size which can be important for the infection outcome (Barber \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2005\u003c/span\u003e; Horn, Liang, and Luong 2023).\u003c/p\u003e \u003cp\u003eWe observed substantial variation in developmental time across generations, but without a consistent trend. The individual variation that we found may be driven by some uncontrolled factors such as subtle differences in parental age that we did not control for.\u003c/p\u003e \u003cp\u003eSome variation between generations was also observed in the microbiota analysis. However, fluctuations in microbial diversity is expected as \u003cem\u003eT. castaneum\u003c/em\u003e microbiota depends on its environment (Agarwal and Agashe 2020). Our results showed that, although overall microbial richness (Chao1 index) was relatively stable under niche and pathogen manipulations, microbial communities in \u003cem\u003eniche-free\u003c/em\u003e regimes seemed to be more consistent, with less variability than \u003cem\u003eniche\u003c/em\u003e regimes. These findings suggest that constructed niches such as flour of \u003cem\u003eT.castaneum\u003c/em\u003e can favour greater variation in microbial community compared to non-constructed niches regardless of infection status. The presence of quinones, which are known to inhibit microbial growth (Armstrong, Spink, and Kahnke 1943; Duarte et al. \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), may have contributed to the increased compositional variation of microbiota observed in larvae of the \u003cem\u003eniche\u003c/em\u003e regimes. By selectively favoring quinone-resistant or less susceptible microbial taxa, the ability to release quinones for niche construction may support greater microbiome diversity. This is consistent with previous findings suggesting that quinones can enhance pathogen defense while sustaining beneficial microbial populations (Yezerski et al. \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2007\u003c/span\u003e; Sawada et al. \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Interestingly, several potentially pathogenic genera\u0026mdash;such as \u003cem\u003eBacillus\u003c/em\u003e, \u003cem\u003eStaphylococcus\u003c/em\u003e, and \u003cem\u003ePseudomonas\u003c/em\u003e\u0026mdash;appeared to be more prevalent in \u003cem\u003eniche-free\u003c/em\u003e regimes (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eD). Moreover, the specific microorganism taxa influenced by the quinones might differ significantly, as it is known that microbes exhibit varying levels of susceptibility to quinones (Yezerski et al. \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2007\u003c/span\u003e; Duarte et al. \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). This variability highlights the complexity of microbial interactions and the potential selective pressures exerted by quinones. Conversely, in individuals from the non-constructed niches regimes where the environment was less controlled, there tended to be a lower variation in larval microbiota, illustrating how environmental control and niche construction can influence microbial community structure.\u003c/p\u003e \u003cp\u003eHost-parasite interactions can influence and be influenced by the microbiome of the host, leading to changes in abundance of specific bacterial taxa (Fredensborg et al. \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Unlike in the western corn rootworm, \u003cem\u003eDiabrotica virgifera\u003c/em\u003e (Chrysomelidae), where microbial diversity was found to be significantly decreased in resistant animals (Paddock et al. \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), in our study, microbial variation was reduced only in larvae of \u003cem\u003eniche-free\u003c/em\u003e regimes. While we observed variation in microbiota composition across evolutionary regimes, this variation did not seem to impair host fitness of \u003cem\u003eT. castaneum\u003c/em\u003e, indicating that functional stability within the microbial community is preserved (Fassarella et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). By altering their microbial environment through stink gland secretions, beetles may reduce the risk of infection by eliminating pathogens, thereby enhancing their survival and reproductive success (Joop et al. \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2014\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn the experimental evolution over 18 generations, pathogen selection led to strong adaptations, with the beetles almost completely losing their susceptibility to the \u003cem\u003eBtt\u003c/em\u003e infection. Via transcriptome analysis, Lo \u0026amp; Schulz et al. (2025) observed clear differences in resistance mechanisms depending on niche construction already after nine generations. While in the presence of a constructed niche the pathogen resistance was achieved mainly through changes in inducible responses, in the regime without the constructed niche resistance correlated with changes in constitutive responses (Lo and Schulz et al. 2025). Does this mean that the external immune response also diverged at a similar rate between the selection regimes? Does the repeated development in conditions with strongly reduced SGS lead to a higher output of this external immune response in adulthood? In other words, can we observe a response to the constant eco-evo feedback created by manipulating the constructed niches? At least in regard to the amount and composition of secreted metabolites the answer appears to be \u0026ldquo;no\u0026rdquo;. We found neither significant differences between the regimes when comparing the amounts of the most abundant metabolites in the secretions, nor was there any distinct differentiation in the metabolite composition of the SGS. While our employed method has its limits in sensitivity, the large sample size we used should have enabled us to also detect minute but consistent differences in secretion composition between the regimes. Thus, it appears that the SGS metabolites and the external immunity they provide are not part of the toolbox involved in resistance evolution. This might be explained by the fact that \u003cem\u003eBtt\u003c/em\u003e infects internally via the midgut after ingestion (Palma et al. \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). The spores only germinate inside the host (Raymond et al. \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2010\u003c/span\u003e), thus more sensitive vegetative cells do not come into direct contact with the host\u0026rsquo;s secretions. Therefore, the adaptation via more \u003cem\u003eBtt\u003c/em\u003e specific immune traits, e.g. upregulation of specific antimicrobial peptide gene expression in the midgut (Baur et al. \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) is likely to be more successful. However, we only analyse changes in metabolites here, but the beetles could also alter their external immune functions by altering the protein composition of the secretions. To test this, proteomic analysis of the SGS or detecting changes in antimicrobial activity via zone of inhibition assays (Lo and Schulz et al. 2025) would be useful as well as detailed analysis of the conditioned flour niche. Finally, we also did not find an effect of evolving in a constructed niche on the SGS profiles, thus the beetles also responded to the less controlled microbial niche caused by the drastically lowered SGSs with altering other traits than their quinone defence. This might also help us to understand the relatively small changes in their associated microbiota. It is also important to note that both SGS profiles and microbiome composition were analysed in F2 offspring. This supports the idea that the observed differences between selection regimes are stable and likely reflect underlying genetic changes rather than transient environmental effects. If these differences are indeed genetically encoded, it is reasonable to assume that similar microbiome patterns would be detectable across different life stages, including both larvae and adults.\u003c/p\u003e \u003cp\u003eOverall, this study provides novel insights into the modulation of host-associated microbiota within complex ecological niches. Our data indicates that a niche constructing trait itself, the SGS, did not change during evolutionary adaptation to a pathogen in the presence vs. absence of a constructed niche, while the larval microbiota showed small but potentially relevant differences between these niche regimes. However, more studies are needed (e.g., on the niche associated microbiome) to conclude whether this has significantly contributed to ecological inheritance.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003ch2\u003eCompeting interest\u003c/h2\u003e \u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eAuthor contribution\u003c/h2\u003e \u003cp\u003eN.K.E.S. and J.K. conceptualized the study. N.K.E.S, H.J., J.F., C.M., and J.K. designed the experiments. N.K.E.S. and L.K.L maintained the selection lines. N.K.E.S., H.J., and J.F. performed the experiments. N.K.E.S., A.K. and J.F. analysed the data. N.K.E.S., A.K., and J.F. wrote the original manuscript draft. All the authors revised the manuscript.\u003c/p\u003e\u003ch2\u003eAcknowledgment\u003c/h2\u003e \u003cp\u003eWe would like to thank Kathrin Br\u0026uuml;ggemann for all the help in the lab, and Paula Mayer, Tobias Pr\u0026uuml;ser and Tobias Gryczan for the help with selection line maintenance. We acknowledge funding by the Deutsche Forschungsgemeinschaft (DFG, German Research 617 Foundation) as part of the SFB TRR 212 (NC\u0026sup3;) \u0026ndash; Project numbers 316099922, 396780003 (to JK) and 396782989 (to CM).\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eAgarwal, Aparna, and Deepa Agashe. 2020. \u0026ldquo;The Red Flour Beetle Tribolium Castaneum: A Model for Host-Microbiome Interactions.\u0026rdquo; Edited by Jeffrey P. Demuth. \u003cem\u003ePLOS ONE\u003c/em\u003e 15 (10): e0239051. https://doi.org/10.1371/journal.pone.0239051.\u003c/li\u003e\n \u003cli\u003eAitchison, J., C. Barcel\u0026oacute;-Vidal, J. A. Mart\u0026iacute;n-Fern\u0026aacute;ndez, and V. Pawlowsky-Glahn. 2000. \u0026ldquo;Logratio Analysis and Compositional Distance.\u0026rdquo; \u003cem\u003eMathematical Geology\u003c/em\u003e 32 (3): 271\u0026ndash;75. https://doi.org/10.1023/A:1007529726302.\u003c/li\u003e\n \u003cli\u003eArmstrong, W. D., Wesley W. Spink, and Jeanne Kahnke. 1943. \u0026ldquo;Antibacterial Effects of Quinones.\u0026lowast;.\u0026rdquo;\u0026nbsp;\u003cem\u003eProceedings of the Society for Experimental Biology and Medicine\u003c/em\u003e 53 (2): 230\u0026ndash;34. https://doi.org/10.3181/00379727-53-14262.\u003c/li\u003e\n \u003cli\u003eBarber, Iain. 2005. \u0026ldquo;Parasites Grow Larger in Faster Growing Fish Hosts.\u0026rdquo; \u003cem\u003eInternational Journal for Parasitology\u003c/em\u003e 35 (2): 137\u0026ndash;43. https://doi.org/10.1016/j.ijpara.2004.11.010.\u003c/li\u003e\n \u003cli\u003eBates, Douglas, Martin M\u0026auml;chler, Ben Bolker, and Steve Walker. 2015. \u0026ldquo;Fitting Linear Mixed-Effects Models Using Lme4.\u0026rdquo; \u003cem\u003eJournal of Statistical Software\u003c/em\u003e 67 (October):1\u0026ndash;48. https://doi.org/10.18637/jss.v067.i01.\u003c/li\u003e\n \u003cli\u003eBates, Douglas, Martin Maechler, and Mikael Jagan. 2000. \u0026ldquo;Matrix: Sparse and Dense Matrix Classes and Methods.\u0026rdquo; https://doi.org/10.32614/CRAN.package.Matrix.\u003c/li\u003e\n \u003cli\u003eBaur, Moritz, Nora K. E. Schulz, Lilo Greune, Zoe M. L\u0026auml;nger, J\u0026uuml;rgen Eirich, Iris Finkemeier, Robert Peu\u0026szlig;, Petra Dersch, and Joachim Kurtz. 2024. \u0026ldquo;Immune Priming in the Insect Gut: A Dynamic Response Revealed by Ultrastructural and Transcriptomic Changes.\u0026rdquo; bioRxiv. https://doi.org/10.1101/2024.11.21.624623.\u003c/li\u003e\n \u003cli\u003eBi, Jie, Lirui Zhang, and Yanping He. 2024. \u0026ldquo;Metabolomics and Microbiome Provide New Insights into Mechanisms of the Variations between the Larva and Adult Stages of the Model Beetle \u003cem\u003eTribolium Castaneum\u003c/em\u003e (Herbst).\u0026rdquo; \u003cem\u003eJournal of Asia-Pacific Entomology\u003c/em\u003e 27 (2): 102241. https://doi.org/10.1016/j.aspen.2024.102241.\u003c/li\u003e\n \u003cli\u003eBreiman, Leo, Adele Cutler, Andy Liaw, and Matthew Wiener. 2002. \u0026ldquo;randomForest: Breiman and Cutlers Random Forests for Classification and Regression.\u0026rdquo; https://doi.org/10.32614/CRAN.package.randomForest.\u003c/li\u003e\n \u003cli\u003eCotter, Sheena C., and R.M. Kilner. 2010. \u0026ldquo;Personal Immunity versus Social Immunity.\u0026rdquo; \u003cem\u003eBehavioral Ecology\u003c/em\u003e 21 (4): 663\u0026ndash;68. https://doi.org/10.1093/beheco/arq070.\u003c/li\u003e\n \u003cli\u003eCotter, Sheena C., Joanne E. Littlefair, Peter J. Grantham, and Rebecca M. Kilner. 2013. \u0026ldquo;A Direct Physiological Trade-off between Personal and Social Immunity.\u0026rdquo; \u003cem\u003eJournal of Animal Ecology\u003c/em\u003e 82 (4): 846\u0026ndash;53. https://doi.org/10.1111/1365-2656.12047.\u003c/li\u003e\n \u003cli\u003eDavyt-Colo, Bel\u0026eacute;n, Juan R. Girotti, Andr\u0026eacute;s Gonz\u0026aacute;lez, and Nicol\u0026aacute;s Pedrini. 2022. \u0026ldquo;Secretion and Detection of Defensive Compounds by the Red Flour Beetle Tribolium Castaneum Interacting with the Insect Pathogenic Fungus Beauveria Bassiana.\u0026rdquo;\u0026nbsp;\u003cem\u003ePathogens (Basel, Switzerland)\u003c/em\u003e 11 (5): 487. https://doi.org/10.3390/pathogens11050487.\u003c/li\u003e\n \u003cli\u003eDen Dool, H. van, and P. Dec. Kratz. 1963. \u0026ldquo;A Generalization of the Retention Index System Including Linear Temperature Programmed Gas\u0026mdash;Liquid Partition Chromatography.\u0026rdquo; \u003cem\u003eJournal of Chromatography A\u003c/em\u003e 11 (January):463\u0026ndash;71. https://doi.org/10.1016/S0021-9673(01)80947-X.\u003c/li\u003e\n \u003cli\u003eDuarte, S\u0026oacute;nia, Ana Magro, Joanna Tom\u0026aacute;s, Carolina Hil\u0026aacute;rio, Ricardo Boavida Ferreira, and Maria Ot\u0026iacute;lia Carvalho. 2022. \u0026ldquo;Antifungal Activity of Benzoquinones Produced by Tribolium Castaneum in Maize-Associated Fungi.\u0026rdquo; \u003cem\u003eInsects\u003c/em\u003e 13 (10): 868. https://doi.org/10.3390/insects13100868.\u003c/li\u003e\n \u003cli\u003eEl-Sayed, Ashraf M. 2024. \u0026ldquo;The Pherobase: Database of Pheromones and Semiochemicals | The World Largest Database of Behavioural Modifying Chemicals.\u0026rdquo; 2024. https://pherobase.com/.\u003c/li\u003e\n \u003cli\u003eFassarella, Marina, Ellen E. Blaak, John Penders, Arjen Nauta, Hauke Smidt, and Erwin G. Zoetendal. 2021. \u0026ldquo;Gut Microbiome Stability and Resilience: Elucidating the Response to Perturbations in Order to Modulate Gut Health.\u0026rdquo;\u0026nbsp;\u003cem\u003eGut\u003c/em\u003e 70 (3): 595\u0026ndash;605. https://doi.org/10.1136/gutjnl-2020-321747.\u003c/li\u003e\n \u003cli\u003eFerro, Kevin, Robert Peu\u0026szlig;, Wentao Yang, Philip Rosenstiel, Hinrich Schulenburg, and Joachim Kurtz. 2019. \u0026ldquo;Experimental Evolution of Immunological Specificity.\u0026rdquo; \u003cem\u003eProceedings of the National Academy of Sciences of the United States of America\u003c/em\u003e 116 (41): 20598\u0026ndash;604. https://doi.org/10.1073/pnas.1904828116.\u003c/li\u003e\n \u003cli\u003eFredensborg, Brian L., Inga Fossdal \u0026iacute; K\u0026aacute;lval\u0026iacute;\u0026eth;, Thor B. Johannesen, C. Rune Stensvold, Henrik V. Nielsen, and Christian M. O. Kapel. 2020. \u0026ldquo;Parasites Modulate the Gut-Microbiome in Insects: A Proof-of-Concept Study.\u0026rdquo; Edited by Simon Russell Clegg. \u003cem\u003ePLOS ONE\u003c/em\u003e 15 (1): e0227561. https://doi.org/10.1371/journal.pone.0227561.\u003c/li\u003e\n \u003cli\u003eFuto, Momir, Sophie A O Armitage, and Joachim Kurtz. 2016. \u0026ldquo;Microbiota Plays a Role in Oral Immune Priming in Tribolium Castaneum.\u0026rdquo; \u003cem\u003eFrontiers in Microbiology\u003c/em\u003e. https://doi.org/10.3389/fmicb.2015.01383.\u003c/li\u003e\n \u003cli\u003eGeib, Scott M., Maria del Mar Jimenez-Gasco, John E. Carlson, Ming Tien, Randa Jabbour, and Kelli Hoover. 2009. \u0026ldquo;Microbial Community Profiling to Investigate Transmission of Bacteria between Life Stages of the Wood-Boring Beetle, Anoplophora Glabripennis.\u0026rdquo; \u003cem\u003eMicrobial Ecology\u003c/em\u003e 58 (1): 199\u0026ndash;211. https://doi.org/10.1007/s00248-009-9501-4.\u003c/li\u003e\n \u003cli\u003eHorn, Collin J., Caroline Liang, and Lien T. Luong. 2023. \u0026ldquo;Parasite Preferences for Large Host Body Size Can Drive Overdispersion in a Fly-Mite Association.\u0026rdquo; \u003cem\u003eInternational Journal for Parasitology\u003c/em\u003e 53 (7): 327\u0026ndash;32. https://doi.org/10.1016/j.ijpara.2023.03.003.\u003c/li\u003e\n \u003cli\u003eJahnes, Benjamin C., Madeline Herrmann, and Zakee L. Sabree. 2019. \u0026ldquo;Conspecific Coprophagy Stimulates Normal Development in a Germ-Free Model Invertebrate.\u0026rdquo; \u003cem\u003ePeerJ\u003c/em\u003e 7:e6914. https://doi.org/10.7717/peerj.6914.\u003c/li\u003e\n \u003cli\u003eJoop, G., O. Roth, P. Schmid-Hempel, and J. Kurtz. 2014. \u0026ldquo;Experimental Evolution of External Immune Defences in the Red Flour Beetle.\u0026rdquo; \u003cem\u003eJournal of Evolutionary Biology\u003c/em\u003e 27 (8): 1562\u0026ndash;71. https://doi.org/10.1111/jeb.12406.\u003c/li\u003e\n \u003cli\u003eK\u0026ouml;rner, Maximilian, Sandra Steiger, and Shantanu P. Shukla. 2023. \u0026ldquo;Microbial Management as a Driver of Parental Care and Family Aggregations in Carrion Feeding Insects.\u0026rdquo; \u003cem\u003eFrontiers in Ecology and Evolution\u003c/em\u003e 11 (August). https://doi.org/10.3389/fevo.2023.1252876.\u003c/li\u003e\n \u003cli\u003eKor\u0026scaron;a, Ana, Lai Ka Lo, Shrey Gandhi, Corinna Bang, and Joachim Kurtz. 2022. \u0026ldquo;Oral Immune Priming Treatment Alters Microbiome Composition in the Red Flour Beetle Tribolium Castaneum.\u0026rdquo;\u0026nbsp;\u003cem\u003eFrontiers in Microbiology\u003c/em\u003e 13. https://www.frontiersin.org/articles/10.3389/fmicb.2022.793143.\u003c/li\u003e\n \u003cli\u003eKov\u0026aacute;ts, E. 1958. \u0026ldquo;Gas-Chromatographische Charakterisierung Organischer Verbindungen. Teil 1: Retentionsindices Aliphatischer Halogenide, Alkohole, Aldehyde Und Ketone.\u0026rdquo; \u003cem\u003eHelvetica Chimica Acta\u003c/em\u003e 41 (7): 1915\u0026ndash;32. https://doi.org/10.1002/hlca.19580410703.\u003c/li\u003e\n \u003cli\u003eLaland, K. N., F. J. Odling-Smee, and M. W. Feldman. 1999. \u0026ldquo;Evolutionary Consequences of Niche Construction and Their Implications for Ecology.\u0026rdquo; \u003cem\u003eProceedings of the National Academy of Sciences\u003c/em\u003e 96 (18): 10242\u0026ndash;47. https://doi.org/10.1073/pnas.96.18.10242.\u003c/li\u003e\n \u003cli\u003eLenth, Russell V. 2017. \u0026ldquo;Emmeans: Estimated Marginal Means, Aka Least-Squares Means.\u0026rdquo; https://doi.org/10.32614/CRAN.package.emmeans.\u003c/li\u003e\n \u003cli\u003eLi, Jianwei, Sabrina Lehmann, Bernhard Wei\u0026szlig;becker, Irene Ojeda Naharros, Stefan Sch\u0026uuml;tz, Gerrit Joop, and Ernst A. Wimmer. 2013. \u0026ldquo;Odoriferous Defensive Stink Gland Transcriptome to Identify Novel Genes Necessary for Quinone Synthesis in the Red Flour Beetle, Tribolium Castaneum.\u0026rdquo; \u003cem\u003ePLOS Genetics\u003c/em\u003e 9 (7): e1003596. https://doi.org/10.1371/journal.pgen.1003596.\u003c/li\u003e\n \u003cli\u003eLo, Lai Ka, Reshma R, Lisa Johanna Tewes, Barbara Milutinović, Caroline M\u0026uuml;ller, and Joachim Kurtz. 2023. \u0026ldquo;Immune Stimulation via Wounding Alters Chemical Profiles of Adult Tribolium Castaneum.\u0026rdquo; \u003cem\u003eJournal of Chemical Ecology\u003c/em\u003e 49 (1): 46\u0026ndash;58. https://doi.org/10.1007/s10886-022-01395-x.\u003c/li\u003e\n \u003cli\u003eLo, Lai Ka, Nora K.E. Schulz, Helle Jensen and Joachim Kurtz. 2025 \u003cem\u003e\u0026quot;\u003c/em\u003eExperimental Removal of Niche construction Alters the Pace and Mechanisms of Resistance Evolution.\u003cem\u003e\u0026quot;\u0026nbsp;\u003c/em\u003ebioRxiv. 2025.03.27.645637; doi: https://doi.org/10.1101/2025.03.27.645637\u003c/li\u003e\n \u003cli\u003eMaino, James L., Paul A. Umina, and Ary A. Hoffmann. 2018. \u0026ldquo;Climate Contributes to the Evolution of Pesticide Resistance.\u0026rdquo; \u003cem\u003eGlobal Ecology and Biogeography\u003c/em\u003e 27 (2): 223\u0026ndash;32. https://doi.org/10.1111/geb.12692.\u003c/li\u003e\n \u003cli\u003eMasri, Leila, Antoine Branca, Anna E. Sheppard, Andrei Papkou, David Laehnemann, Patrick S. Guenther, Swantje Prahl, et al. 2015. \u0026ldquo;Host\u0026ndash;Pathogen Coevolution: The Selective Advantage of Bacillus Thuringiensis Virulence and Its Cry Toxin Genes.\u0026rdquo; \u003cem\u003ePLoS Biology\u003c/em\u003e. https://doi.org/10.1371/journal.pbio.1002169.\u003c/li\u003e\n \u003cli\u003eMcMurdie, Paul J., and Susan Holmes. 2013. \u0026ldquo;Phyloseq: An R Package for Reproducible Interactive Analysis and Graphics of Microbiome Census Data.\u0026rdquo; Edited by Michael Watson. \u003cem\u003ePLoS ONE\u003c/em\u003e 8 (4): e61217. https://doi.org/10.1371/journal.pone.0061217.\u003c/li\u003e\n \u003cli\u003eM\u0026uuml;ller, Caroline, Barbara A. Caspers, J\u0026uuml;rgen Gadau, and Sylvia Kaiser. 2020. \u0026ldquo;The Power of Infochemicals in Mediating Individualized Niches.\u0026rdquo; \u003cem\u003eTrends in Ecology \u0026amp; Evolution\u003c/em\u003e 35 (11): 981\u0026ndash;89. https://doi.org/10.1016/j.tree.2020.07.001.\u003c/li\u003e\n \u003cli\u003eNikodemova, Maria, Elizabeth A. Holzhausen, Courtney L. Deblois, Jodi H. Barnet, Paul E. Peppard, Garret Suen, and Kristen M. Malecki. 2023. \u0026ldquo;The Effect of Low-Abundance OTU Filtering Methods on the Reliability and Variability of Microbial Composition Assessed by 16S rRNA Amplicon Sequencing.\u0026rdquo; \u003cem\u003eFrontiers in Cellular and Infection Microbiology\u003c/em\u003e 13 (June). https://doi.org/10.3389/fcimb.2023.1165295.\u003c/li\u003e\n \u003cli\u003eOdling-Smee, F. John, Kevin N. Laland, and Marcus W. Feldman. 2013. \u003cem\u003eNiche Construction: The Neglected Process in Evolution\u003c/em\u003e. Princeton University Press. https://doi.org/10.1515/9781400847266.\u003c/li\u003e\n \u003cli\u003eOksanen, Jari, Gavin L. Simpson, F. Guillaume Blanchet, Roeland Kindt, Pierre Legendre, Peter R. Minchin, R.B. O\u0026rsquo;Hara, et al. 2001. \u0026ldquo;Vegan: Community Ecology Package.\u0026rdquo; https://doi.org/10.32614/CRAN.package.vegan.\u003c/li\u003e\n \u003cli\u003eOttensmann, Meinolf, Martin A. Stoffel, Hazel J. Nichols, and Joseph I. Hoffman. 2018. \u0026ldquo;GCalignR: An R Package for Aligning Gas-Chromatography Data for Ecological and Evolutionary Studies.\u0026rdquo; \u003cem\u003ePLOS ONE\u003c/em\u003e 13 (6): e0198311. https://doi.org/10.1371/journal.pone.0198311.\u003c/li\u003e\n \u003cli\u003eOtti, Oliver, Simon Tragust, and Heike Feldhaar. 2014. \u0026ldquo;Unifying External and Internal Immune Defences.\u0026rdquo; \u003cem\u003eTrends in Ecology \u0026amp; Evolution\u003c/em\u003e 29 (11): 625\u0026ndash;34. https://doi.org/10.1016/j.tree.2014.09.002.\u003c/li\u003e\n \u003cli\u003ePaddock, Kyle J., Adriano E. Pereira, Deborah L. Finke, Aaron C. Ericsson, Bruce E. Hibbard, and Kent S. Shelby. 2021. \u0026ldquo;Host Resistance to Bacillus Thuringiensis Is Linked to Altered Bacterial Community within a Specialist Insect Herbivore.\u0026rdquo; \u003cem\u003eMolecular Ecology\u003c/em\u003e 30 (21): 5438\u0026ndash;53. https://doi.org/10.1111/mec.15875.\u003c/li\u003e\n \u003cli\u003ePalma, Leopoldo, Delia Mu\u0026ntilde;oz, Colin Berry, Jes\u0026uacute;s Murillo, Primitivo Caballero, and Primitivo Caballero. 2014. \u0026ldquo;Bacillus Thuringiensis Toxins: An Overview of Their Biocidal Activity.\u0026rdquo;\u0026nbsp;\u003cem\u003eToxins\u003c/em\u003e 6 (12): 3296\u0026ndash;3325. https://doi.org/10.3390/toxins6123296.\u003c/li\u003e\n \u003cli\u003eParadis, Emmanuel, and Klaus Schliep. 2019. \u0026ldquo;Ape 5.0: An Environment for Modern Phylogenetics and Evolutionary Analyses in R.\u0026rdquo; \u003cem\u003eBioinformatics\u003c/em\u003e 35 (3): 526\u0026ndash;28. https://doi.org/10.1093/bioinformatics/bty633.\u003c/li\u003e\n \u003cli\u003eR Core Team. 2024. \u0026ldquo;R: A Language and Environment for Statistical Computing.\u0026rdquo; Vienna, Austria: R Foundation for Statistical Computing. https://www.R-project.org/.\u003c/li\u003e\n \u003cli\u003eRaymond, Ben, Paul R. Johnston, Christina Nielsen-LeRoux, Didier Lereclus, and Neil Crickmore. 2010. \u0026ldquo;Bacillus Thuringiensis: An Impotent Pathogen?\u0026rdquo; \u003cem\u003eTrends in Microbiology\u003c/em\u003e 18 (5): 189\u0026ndash;94. https://doi.org/10.1016/j.tim.2010.02.006.\u003c/li\u003e\n \u003cli\u003eRipley, Brian, and Bill Venables. 2009. \u0026ldquo;MASS: Support Functions and Datasets for Venables and Ripley\u0026rsquo;s MASS.\u0026rdquo; https://doi.org/10.32614/CRAN.package.MASS.\u003c/li\u003e\n \u003cli\u003eRoth, Louis M., and Ruth B. Howland. 1941. \u0026ldquo;Studies on the Gaseous Secretion of Tribolium Confusum Duval I. Abnormalities Produced in Tribolium Confusum Duval by Exposure to a Secretion given off by the Adults.\u0026rdquo; \u003cem\u003eAnnals of the Entomological Society of America\u003c/em\u003e 34 (1): 151\u0026ndash;75. https://doi.org/10.1093/aesa/34.1.151.\u003c/li\u003e\n \u003cli\u003eRStudio Team. 2023. \u0026ldquo;RStudio: Integrated Development for R.\u0026rdquo; PBC, Boston, MA URL: RStudio. www.rstudio.com.\u003c/li\u003e\n \u003cli\u003eSawada, Mitsuki, Takuma Sano, Kento Hanakawa, Patchara Sirasoonthorn, Takao Oi, and Ken Miura. 2020. \u0026ldquo;Benzoquinone Synthesis-Related Genes of Tribolium Castaneum Confer the Robust Antifungal Host Defense to the Adult Beetles through the Inhibition of Conidial Germination on the Body Surface.\u0026rdquo; \u003cem\u003eJournal of Invertebrate Pathology\u003c/em\u003e 169 (January):107298. https://doi.org/10.1016/j.jip.2019.107298.\u003c/li\u003e\n \u003cli\u003eSchwenke, Robin A., Brian P. Lazzaro, and Mariana F. Wolfner. 2016. \u0026ldquo;Reproduction-Immunity Trade-Offs in Insects.\u0026rdquo; \u003cem\u003eAnnual Review of Entomology\u003c/em\u003e 61:239\u0026ndash;56. https://doi.org/10.1146/annurev-ento-010715-023924.\u003c/li\u003e\n \u003cli\u003eTherneau, Terry M., Thomas Lumley (original S.-\u0026gt;R port and R. maintainer until 2009), Atkinson Elizabeth, and Crowson Cynthia. 2024. \u0026ldquo;Survival: Survival Analysis.\u0026rdquo; https://cran.r-project.org/web/packages/survival/index.html.\u003c/li\u003e\n \u003cli\u003eTherneau, Terry M., Patricia M. Grambsch, and V. Shane Pankratz. 2003. \u0026ldquo;Penalized Survival Models and Frailty.\u0026rdquo; \u003cem\u003eJournal of Computational and Graphical Statistics\u003c/em\u003e 12 (1): 156\u0026ndash;75. https://doi.org/10.1198/1061860031365.\u003c/li\u003e\n \u003cli\u003eVera-Ponce de Le\u0026oacute;n, Arturo, Benjamin C. Jahnes, Alejandro Otero-Bravo, and Zakee L. Sabree. 2021. \u0026ldquo;Microbiota Perturbation or Elimination Can Inhibit Normal Development and Elicit a Starvation-Like Response in an Omnivorous Model Invertebrate.\u0026rdquo; \u003cem\u003emSystems\u003c/em\u003e 6 (4): e0080221. https://doi.org/10.1128/mSystems.00802-21.\u003c/li\u003e\n \u003cli\u003eWang, Zhengyan, Haisheng Hu, Zhenzhen Chang, Shan Zhang, and Yujie Lu. 2023. \u0026ldquo;Molecular Mechanism of Enterococcus Faecalis-Induced Phosphine Sensitivity in Tribolium Castaneum (Coleoptera: Tenebrionidae).\u0026rdquo; \u003cem\u003eJournal of Economic Entomology\u003c/em\u003e 116 (5): 1820\u0026ndash;29. https://doi.org/10.1093/jee/toad171.\u003c/li\u003e\n \u003cli\u003eWickham, Hadley, Winston Chang, Lionel Henry, Thomas Lin Pedersen, Kohske Takahashi, Claus Wilke, Kara Woo, et al. 2024. \u0026ldquo;Ggplot2: Create Elegant Data Visualisations Using the Grammar of Graphics.\u0026rdquo; https://cran.r-project.org/web/packages/ggplot2/index.html.\u003c/li\u003e\n \u003cli\u003eYew, Joanne Y., and Henry Chung. 2015. \u0026ldquo;Insect Pheromones: An Overview of Function, Form, and Discovery.\u0026rdquo; \u003cem\u003eProgress in Lipid Research\u003c/em\u003e 59 (July):88\u0026ndash;105. https://doi.org/10.1016/j.plipres.2015.06.001.\u003c/li\u003e\n \u003cli\u003eYezerski, Ann, Carla Ciccone, Jill Rozitski, and Beth Volingavage. 2007. \u0026ldquo;The Effects of a Naturally Produced Benzoquinone on Microbes Common to Flour.\u0026rdquo; \u003cem\u003eJournal of Chemical Ecology\u003c/em\u003e 33 (6): 1217\u0026ndash;25. https://doi.org/10.1007/s10886-007-9293-2.\u003c/li\u003e\n \u003cli\u003eZhang, Husen, Joshua B. Sparks, Saikumar V. Karyala, Robert Settlage, and Xin M. Luo. 2015. \u0026ldquo;Host Adaptive Immunity Alters Gut Microbiota.\u0026rdquo; \u003cem\u003eISME Journal\u003c/em\u003e 9 (3): 770\u0026ndash;81. https://doi.org/10.1038/ismej.2014.165.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[{"identity":"a40a4889-d1d7-4cf4-8ec5-255dbfd44749","identifier":"10.13039/501100001659","name":"Deutsche Forschungsgemeinschaft","awardNumber":"316099922, 396780003 (to JK), and 396782989 (to CM)","order_by":0}],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"University of Münster","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"niche construction, microbiota, flour beetles, external immunity, quinones, experimental evolution","lastPublishedDoi":"10.21203/rs.3.rs-6353988/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6353988/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThrough niche construction, organisms actively shape their environment, thereby influencing their evolutionary trajectories via ecological inheritance. Red flour beetles (\u003cem\u003eTribolium castaneum\u003c/em\u003e) achieve niche construction through secretion of antimicrobial compounds from the stink glands. It has recently been demonstrated that the experimental removal of niche construction using RNAi of a key gene needed to produce stink gland secretions altered the pace and mechanisms of resistance adaptation to the bacterial entomopathogen \u003cem\u003eBacillus thuringiensis\u003c/em\u003e within nine host generations. However, it is unknown whether the microbiome and secretions produced by beetles undergo changes during experimental evolution. We continued the evolution experiment with an additional nine generations of selection. We found that host resistance continued to increase in selection regimes with pathogen exposure, whereas host development and fecundity remained stable, thereby confirming our previous findings. We then profiled larvae-associated microbiota in generations 12 and 15 via 16S rRNA sequencing and measured the stink gland secretion profiles of adults via gas chromatography-flame ionization detection in generation 18. While adaptation to the pathogen did not affect the microbiota, lines evolving with the possibility to construct their niches showed increased microbial diversification, and chemical secretion profiles did not change in either of the selection regimes. Together, our results highlight the role of niche construction in shaping host\u0026ndash;microbe interactions. These effects seemed to be independent of any microevolutionary changes in the secretions as a niche-constructing trait.\u003c/p\u003e","manuscriptTitle":"Resistance evolution driven by niche construction – the role of microbiota and beetle secretions","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-04-02 05:33:40","doi":"10.21203/rs.3.rs-6353988/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"3f3b6ce5-8e27-4733-ae2c-e8c4901a017f","owner":[],"postedDate":"April 2nd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":46526283,"name":"Evolutionary Biology"}],"tags":[],"updatedAt":"2025-04-02T05:33:40+00:00","versionOfRecord":[],"versionCreatedAt":"2025-04-02 05:33:40","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6353988","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6353988","identity":"rs-6353988","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}