You are here:
SOURCE AGGREGATION IN STABLE ISOTOPE MIXING MODELS: LUMP IT OR LEAVE IT?
Citation:
Phillips, D L., J W. Gregg, AND S. D. Newsome. SOURCE AGGREGATION IN STABLE ISOTOPE MIXING MODELS: LUMP IT OR LEAVE IT? Presented at 4th International Conference on Applications of Stable Isotope Techniques to Ecological Studies, Wellington, New Zealand, April 19-23, 2004.
Description:
A common situation when stable isotope mixing models are used to estimate source contributions to a mixture is that there are too many sources to allow a unique solution. To resolve this problem one option is to combine sources with similar signatures such that the number of sources is small enough to provide a unique solution. However, this will tend to increase the uncertainty of the combined end-member and consequently the source contribution estimates, an effect that can be quantified using the IsoError model described by Phillips & Gregg (2001). Common and reasonable practices are to consider lumping only when the isotopic signatures of clustered sources are not significantly different, and when sources are related such that the combined source group has some functional significance. For example, lumping several species within a trophic guild will allow more interpretable results in a dietary analysis than lumping disparate food sources (elephants and amoebae), even if they have similar isotopic signatures. As an alternative to lumping sources before a mixing analysis, the IsoSource mixing model (Phillips and Gregg 2003) can be used to find all feasible solutions of source contributions that are consistent with isotopic mass balance. While the range of feasible contributions for each source can often be quite broad, functionally related groups of sources can be combined post hoc, producing a range of solutions for the aggregate source that may be considerably narrower than that of each individual source. A paleohuman dietary analysis example is given to illustrate this point, which involves a terrestrial meat food source, a combination of three terrestrial plant foods, and a combination of three marine foods. In this case, post hoc aggregation of sources allowed strong conclusions about temporal shifts in marine vs. terrestrial diets that would not have otherwise been discerned.