These results suggest that stable isotope analysis has potential as a tool for distinguishing between tortoises of different origin, particularly when testing that batches of tortoises are consistent with paperwork. For example, if a consignment of tortoises is imported with paperwork claiming that all specimens are captive bred and from the same place, there should not be systematic differences between the signatures of individuals.
However, this does not address the problem of assessing the origin of individual tortoises. It is true that the smooth (likely wild) tortoises in my sample have a consistently higher Nitrogen signature than the pyramided (likely ranched) tortoises. However, it may be that some ranched populations also have a high Nitrogen signature, or that some wild populations have a very low signature.
In order to assess origin of an individual it is therefore necessary to look at something other than actual signature. Degree of change over time seems to show some promise here. Tortoises have relatively small ranges and wild food plants within their range should not show great changes in signature form year to year. On the other hand tortoises in captivity are fed a range of foods and supplements that are likely to change from year to year. We may, therefore, expect to see a greater change in signature over time in captive tortoises. To this end I have tested whether there is a difference between smooth and pyramided tortoises in degree of change over time and have plotted this below (figure 4).
Looking at these plots it appears that the carbon signature of pyramided tortoises does in fact change to a greater degree than for smooth tortoises. I tested this by calculating the mean change over time for each individual and comparing these values using Welch's t test. It turns out there is a significant difference between smooth and pyramided tortoises. In fact there is no overlap in amount of change over time for the two groups; this means that every pyramided tortoise changes more over time than every smooth tortoise. The way I tested this is very crude but given the tiny number of smooth tortoises I didn't have enough power to do much else.
In the next few weeks I will be analysing several more tortoises from the same batch as those above, and will also be running samples for Hydrogen analysis.