As a critical renewable energy source, biomass has substantial potential for reducing carbon emissions and promoting sustainable development. However, its long-term storage in large volumes for modern bioenergy poses challenges due to inherent self-heating driven by exothermic microbial, physical and chemical processes, which can lead to spontaneous ignition and severe fire incidents. While physical and chemical aspects of self-heating in biomass piles greatly benefit from extensive research and experience borrowed from coal pile self-heating, microbial activities in biomass self-heating have received less attention. To address this knowledge gap, we conduct a series of comprehensive studies to gain a deeper understanding of microbial activities in biomass self-heating, aiming to develop a numerical model for reliably predicting biomass self-heating and minimizing fire risks in stored biomass. This paper presents the results of our experimental investigation into microbial respiration during biomass storage. Six commonly used biomass feedstocks with varying initial moisture content are subjected to different storage conditions over a seven-day testing period, during which the main metabolic byproduct (CO2) released from microbial degradation of the biomass is continuously measured. Our findings reveal that an increase in feedstock moisture content significantly enhances microbial activity when the initial moisture content is below 136 %. However, further increases in initial moisture content do not notably enhance, and in some cases, may even weaken microbial activity. Additionally, a temperature range of 25 °C to 45 °C is identified as conducive to rapid microbial decomposition for different biomass materials, and deviation from this range results in a significant reduction in microbial activity. Furthermore, under identical temperature and initial moisture content conditions, the order of carbon dioxide evolution rate is as follows: corn stalk > soybean hull > rice straw > wheat straw > cotton stalk > pepper stalk. Our respiration intensity test results also create a fundamental experimental database for model development and validation, and contribute to the broader understanding of biomass storage and its associated challenges.