A comparison of SIMM assumptions and features among commonly used SIMMs.

Four mixing model assumptions (italics) commonly violated when estimating the proportional dietary contribution of sources to the diets of animals, and the model feature that addresses each violated assumption. A list of other features included in SIMMs and their definitions. X denotes the model addresses the assumption or includes the feature and Y indicates the feature is not explicitly included (e.g., model may account for error using an arbitrary tolerance measure). MCMC (Markov chain Monte Carlo), SIR (sequential importance resampling), and ML (maximum likelihood) denotes sampling method used when estimating parameters.aX denotes that the model provides solutions when sources exceed n+1, but solutions are not comparable to other models (i.e., output lists ranges of potential solutions, not parameter estimates).bX indicates Ward et al. (35) was the first study to use this approach. However, this model (35) has recently been introduced; therefore, it has not been commonly used.