J. Lancaster et S. Waldron, Stable isotope values of lotic invertebrates: Sources of variation, experimental design, and statistical interpretation, LIMN OCEAN, 46(3), 2001, pp. 723-730
In a subset of a stream food web, whole-body isotope values of delta C-13 a
nd delta N-15 were measured for eight populations of lotic invertebrates. O
bserved isotopic differences among species corresponded broadly to their tr
ophic status, as also revealed by gut content analysis, but with some excep
tions. Species within a guild of grazer/scraper mayflies differed significa
ntly in delta C-13; a predatory caddisfly (Rhyacophila dorsalis) and a coll
ector/gatherer stonefly (Leuctra inermis) had statistically indistinguishab
le values of delta C-13 and delta N-15. Th, variation associated with the m
ean isotope value of each population was partitioned into the variation amo
ng individuals and the variation that arises from analysis by isotope ratio
mass spectrometry. For some taxa, within-population variance was lower tha
n or equal to the variance attributable to the measurement error of the mas
s spectrometer. The highest but conservative estimate of within-population
variation was a mean coefficient of variation of 11% for delta N-15 in a pr
edator R. dorsalis. The minimum detectable difference between two populatio
ns was negatively associated with the number of replicate samples and the n
umber of individual animals combined in each replicate. The optimum number
of replicate samples, therefore, varies depending on the hypotheses of inte
rest.