Hi Elena. You are right. One of the reasons has to do with the possible and yet unknown effect of enriched diet on fractionation processes. Basically, we are not sure how would react the routing of molecules when facing unconventional concentration of a certain isotope. But I have some recollection of studies using either enriched diet or diets with weird signature so that there is a huge difference between the diet in situ and in the lab (check some of my papers and references therein).

Personally, I would definitely call for more lab experiments, including using

enriched diets.

Diet-switch experiments are not a method of choice to determine discrimination but for the quantification of isotope turnover (e.g. number of pools, pool half-life). Some of the pools that are turned over have a very long half-life. In consequence, a constant diet has to be maintained for a very very long time after a diet switch, until all the body pools have equilibrated with the new diet and proper discrimination can be determined.

This is why the discrimination should best be obtained under conditions of long-lasting constant diet but not with a diet-switch experiment. A long-lasting constant diet, however, is difficult to provide especially for large animals. In reality, this condition can never be fully met (e.g. due to seasonal changes in diet or because of the changing nutritional needs of an animal from birth to death).

The contribution of slow body pools that have not fully been equilibrated with the present diet will always cause a negative correlation between diet and tissue isotope composition. Such relations are is frequently found and indicate that the condition of equilibration had not been met.

Some additional information...

Hi Elena,

when you think about mass-dependent fractionation factors on the natural abundance level (and this is what we are normally interested in), the different masses do not lead to strong isotope discrimination. That´s why we use in practise delta values and isotope enrichment factors. Therefore, we need highly precise measurements, that can be done with the Nier-type mass spectrometers (speaking about light isotopes) and using appropriate standards to calibrate the measurements to the mass scale.

This brings us to the problem with isotopically enriched material: First of all, most mass spectrometers do not handle routinely high abundances of the normally minor isotopes in high precision. In addition, we are lacking well known enriched standards. isotope mixing can easily be followed, but fractionation factors...

By the way: Regarding the report of isotope abundances, I suggest to look into a nice paper:

Brand, W.A., and Coplen, T.B., 2012, Stable isotope deltas: tiny, yet robust signatures in nature, Isotopes in Environmental and Health Studies, v. 48, n. 3, p. 393-409.

Hope that helps. Best, Michael

Thank you, Stanislas and Karl. Let's say we have a very fast growing animal (Daphnia) kept in culture with algae growing in a chemostat (=virtually invariable isotopic composition, no seasonality) for many-many generations. When studying effects of various biotic and abiotic factors on discrimination (and turnover, of course), one would need to create an isotopically distinct diet and apply a diet-switch approach. This is basically what we did but went wrong in our calculations with isotope labelling of the algae. As a result, the diet was very much outside of the natural range of both 13C and 15N. However, the discrimination factors were outrageous and I am trying to understand what would be the mechanism for this huge diet-body shifts and how relevant this is for interpretation of the experimental results on growth and metabolism contributions to the turnover.

Thank you, Michael. The review is indeed very nice!

Yes, I am aware of the problems with measurements but it is a bit different story. If we could measure with a perfect precision highly enriched samples, what would be the mechanisms of really large values for diet-tissue shift in for example d15N within the same experiment? By really large I mean +200 or -200... with excellent repeatability. This iis what we can induce using the same enriched diet (d15N ~1000)...

A long-lasting constant diet, however, is difficult to provide especially for large animals. A constant diet has to be maintained for a very long time after a diet switch, until all the body pools have equilibrated with the new diet and proper discrimination can be determined. In reality, this condition can never be fully met. This brings us to the problem with isotopically enriched material. We need highly precise measurements.

Hi Elena, your problem sounds really interesting. If I understood right, we can exclude the typical errors leading to high shifts like inhomogeneous labeling of the diet or incomplete turnover of body tissues. At the moment I cannot see any reason for this strange finding. I would like to think longer about what could be the underlying mechanism. Could you send me more information about your methods (organisms, method of labeling, feeding, calculation…). The reason must be somewhere there. My email address is [email protected].

One problem could result from the chitin skeleton. The chitin nitrogen must not necessarily originate from the protein of the diet and can thus differ a lot. In case that you find a large enrichment in the body, this must be balanced by a large depletion in the excrements. Is it possible to collect and analyze the excrements. This should give a hint on the mechanisms (digestibility, heterogenity in the diet …)

Thank you, I will try to put together some explanatory text and mail it.

Yes, in this case, the diet labelling was homogenious and animals grew in size 3- to 7-fold (depending on treatment). Of course, there are different pools, like in any other organsim, but since the growth is very rapid, the differences between these pools are indistinguishable (at least with daily sampling). Chitin% in these Daphnia does not exceed 4% of dry weight, and, therefore regardless of its composition (it is usually negative in small crustaceans and built using recycled N-pool), its contribution cannot explain the observed values. Here is a figure showing discrimination for 2 enriched diets that differ in their 13C and 15N composition as a function of growth rate.