Composite virulence: useful metric or conceptual trap?

Luis M. Silva , Tiago G. Zeferino

Virulence, the harm an infection causes to its host, is a cornerstone concept in ecology and evolution, yet it remains difficult to quantify because infection impact is multidimensional, dynamic, and context-dependent. Infections can reduce host performance through multiple, partially redundant routes (including mortality, fecundity loss, behavioural impairment, and physiological disruption), whose relative importance depends on the host's life history and ecological conditions. Composite virulence measures have emerged to address this complexity by combining information across traits, but they can embed strong assumptions: component choice may reflect availability rather than biological relevance, correlated traits can be double-counted, and combining parasite drivers or host-state mediators with life-history outcomes can obscure interpretation and introduce circularity. Here we synthesize the conceptual foundations of composite virulence and outline a biologically grounded workflow for building interpretable composites. We treat composite virulence as a measurement strategy anchored to an explicit biological question, with trait selection conducted to minimize redundancy while preserving timing and context-dependency. Used carefully, composite approaches can clarify which routes to harm dominate across hosts, parasites, and environments and thereby strengthen evolutionary inference; used casually, they risk becoming a conceptual trap that yields clean rankings while concealing mechanisms.