Observational field studies are not appropriate tests of consumer stoichiometric homeostasis

Abstract

Stoichiometric homeostasis is a key concept of ecological stoichiometry theory that describes organisms' ability to maintain internal elemental composition given variable resource elemental composition. Observational field studies are used frequently to examine homeostasis, and >1/2 of studies assessing homeostasis in the last 15 y were based on field data. We explored whether field surveys are appropriate to assess consumer homeostasis. We used examples from benthic macroinvertebrates to argue that most field data are inappropriate to assess homeostasis because investigators using field studies: 1) face difficulty accurately characterizing consumer and resource stoichiometry, 2) rarely use sufficient range and replication across discrete levels of resource stoichiometry, and 3) are challenged by great variation of consumer stoichiometry resulting from confounding factors (body size, ontogeny, sex, and others) across study sites. We used a simple modeling approach to show how ontogeny and tissuespecific turnover time can change consumer stoichiometry over time to confound homeostasis assessments. This situation is a pitfall in many field studies in which instantaneous measures of body stoichiometry are used without control for body size. We used power analyses combined with literature meta-analysis to show that most field studies possess limited or no statistical power to test homeostasis because of their low sample sizes and greater variation in consumer stoichiometry compared to laboratory assessments. Laboratory experiments better control consumer and resource stoichiometry, decrease effects of body size, ontogeny, and other confounding factors to reduce model variation, and increase replication for robust tests of homeostasis. Our concluding recommendations for appropriate design and interpretation of laboratory experiments will help scientists resolve factors driving consumer stoichiometry in the field.