The soil decomposer community is a primary driver of carbon cycling in forest ecosystems. Understanding the processes that regulate this community is critical to our understanding of the global carbon cycle and fungal mediated impact on climate change. Inadequate statistical strength in traditional soil food web studies has limited our capacity to disentangle the cascading effect of top-level predators on the composition of complex fungal communities. We hypothesize that machine learning can help with this complex problem. This paper examines the opportunities for machine learning in this domain, presents initial results from such analysis, identifies challenges encountered with initial effort, and charts a path forward.