Traditionally, philosophy uses the discrimination of self and non-self to define individuality with the immune system performing this discrimination. In the evolutionary field of biology this distinction is not that simple. Nowadays it is becoming more and more apparent that individuals can no longer be considered as ‘lone isolated islands’ in the ‘environmental sea’. All kinds of eukaryotic taxa harbour their own microbiota consisting of bacteria, archaea, fungi, protozoa and viruses and they are tolerated by the host’s immune system because of their manifold beneficial functions on, for example, host nutrition, detoxification, development, fecundity or pathogen protection. However, not only the beneficial microbiome, but also the host’s nutrition can strongly affect its physiology and its ability to combat pathogen infections. Microbiome and host form a unit – the holobiome. Notably, even though we gained insights on either the function of the microbiome or of the nutrition on host immune defence in diverse separate studies we still poorly understand how they act together in particular organisms. An insect model system to study these interactions are cockroaches. This is because, 1) they are omnivorous generalists, which makes them easily accessible for nutritional studies; 2) they harbour a diverse microbiota which can be manipulated through sterilization methods; and 3) they feature effective strategies to combat pathogens since they are frequently exposed to a rich antigenic environment due to their lifestyle. First, in Chapter I I investigated the nutritional dependencies of immunity in the cockroach system by performing food choice experiments using the cockroach species Blatta orientalis upon exposure to the opportunistic Gram-negative bacterial insect pathogen Pseudomonas entomophila. I could show that depending on the strength of infection B. orientalis males reduce their overall nutrient intake and increase the protein to carbohydrate ratio being consumed. Interestingly, these behavioural shifts do not boost the insect’s immunity as indicated by the examination of the hemolymph’s antimicrobial activity, the abundance of immune proteins in the hemolypmph or the general host survival. This lack of benefits for the host highlights a possible decoupling of dietary macronutrient regulation from immunity in these invasive animals with the possibility that anorexia, in general, might be a more powerful tool if diet quality is highly unpredictable for generalist species. In Chapter II I evaluated two different approaches for the development of a germ-free Blatella germanica cockroach breeding system which forms the basis of any study dealing with the function of the cockroach microbiome. While one of these methods uses peracetic acid, the other one uses a combination of peracetic acid and sodium hypochlorite to surface-sterilize cockroach oothecas to deprive the hatchlings from their natural microbiota. These treatments should leave them only with their obligate symbiont Blattabacterium sp., which supplies essential vitamins and is required for the development into fecundant adults. I tested the success of those techniques by plating adult individuals on Lysogeny-broth-agar and by using state of art 16S metabarcoding. It turned out that both methods performed quite poorly leading to individuals which can be considered as germ-free in 40 % of all cases. I therefore developed our own method by combining sequential ootheca surface sterilizations with peracetic acid and sodium hypochlorite followed by a treatment of freshly hatched nymphs with the antibiotics rifampicin and gentamicin which significantly improved its effectiveness resulting in germ-free adult cockroaches to 99 % of all cases. In addition, I used our germ-free cockroach system for an early study on the impact of the absence of an intact microbiome on developmental time. I could show that B. germanica cockroaches deprived of their natural microbiota needed approximately 35 days longer from the day of hatching to the day molting into adults than their conventional counterparts, which already grants a small glimpse on the strong impacts of the microbiome on the host physiology and its overall performance. In Chapter III I analysed the transcriptome of germ-free and conventional B. germanica males and followed their survival upon P. entomophila systemic infection to gain further insights on the influence of the cockroach microbiome on host traits. The basis of our gene identification were two published genomes either the one by Harrison et al. (2018) or the one by He (2018). Depending on the reference genome used for the analyses small differences existed. When the Harrison et al. genome was used 25451 putative genes were identified and 184 of those including 19 immune-related genes were significantly different expressed between conventional and germ-free cockroaches. When the He genome was used 111778 putative genes were identified and 1082 of those including 30 immune-related genes were significantly different expressed between conventional and germ-free cockroaches. Immune-related genes which were significantly expressed between germ-free and conventional cockroaches identified with both reference genomes included hemolymph lps-binding protein related genes which were mostly upregulated in germ-free individuals because of their role in trapping the Blattabacterium sp. endosymbiont and tenecin-1 genes, a transferrin, a caspase-1, a alpha-1-macroglobulin, a lysozym c-1 as well as a catalase found to be upregulated in conventional individuals. All the latter ones contribute to the recognition and the suppression of microbial life to maintain a stable host microbiota. This regulation of gene expression by the microbiome might also assist the host in combating infections as indicated by the significantly higher survival of conventional cockroaches infected with P. entomophila. In conclusion I was able to show, that host biology is heavily shaped by microbial life. Those microbes can be either invading pathogens or commensal or beneficial microbiota. In all cases they radically alter host behaviour, development and/or physiology. While pathogens only harm their hosts, the microbiome promotes host phenotypes like development or immune competence. Therefore, pathogens do not only interact with their hosts but also with its microbiota. Since this fact became only apparent within the last few years more research is needed to reveal all its aspects. A stable foundation for such future work is paved by my recently established germ-free B. germanica breeding system. In this framework it will be particularly important and likewise exciting to perform refaunation experiments with single microbial taxa followed by infections with different pathogens and further transcriptomic analyses to uncover their special tasks in this broad network of interactions.