The gut microbiome plays a vital role in human health. Disturbances of this microbial system is associated with diseases such as obesity and inflammatory bowel disease. In populations of microbial species, many organisms partake in byproduct cross-feeding interactions, where byproducts from one organism are consumed by other microbes. Using the digital evolution software Avida, I studied the effect of recurring perturbations and imperfect resource conversion on the evolution of byproduct cross-feeding chains in digital communities. To investigate the effect of perturbation and conversion rate on digital organisms, I evolved digital communities for 200,000 updates in an unperturbed environment that could hold 50 different resource types, each produced as a byproduct of consuming another resource. At 200,000 updates, 50 or 60 % of all organisms were removed at various intervals during periods of different lengths, with a conversion rate less than 100 % between resources in the byproduct chain. I found that 0.9 conversion rate caused communities to evolve longer cross-feeding chains. A conversion rate of 0.5 resulted in communities with much shorter chains, more similar in length to byproduct chains in the human gut. Perturbation events seem to affect chain length only under certain conditions when energy is lost between resources, for example when 60 % of all organisms were removed every 50th update on average. It appears that conversion loss makes digital communities more robust against the effects of perturbations, and that it might protect these communities from going extinct.