The Project
Preliminary results

As I had hoped, my preliminary results suggest that there is no significant difference between scutes (the hexagonal sections that make up the shell), and no significant difference between the top (carapace) and underside (plastron) of the shell.  This means that I will be able to pool material from multiple scutes to get a better sample weight.   

So: no significant variation between the scutes.  There is however a fair amount of variation between the different ages of growth, particularly for nitrogen: the tissue the tortoise was born with (the juvenile scute) is more enriched in Nitrogen than other growth layers (figure 1).  Nitrogen becomes enriched in the heavier isotope as trophic level increases; a predator will be more enriched in 15N than it's prey.  These results may therefore be because Horsfield's tortoises are capital breeders, which means they use stored resources for reproduction.  The tissues of the offspring are produced from metabolised fat and are thus effectively a trophic level higher than the tissues of the mother.  

Figure 1: Isotope signatures across the scute. Left: Nitrogen, right: Carbon

Interestingly, there is also a significant change in the carbon signature across the scute (see graphs above), which could be due to a change in diet or location, but I don't have enough information about the tortoises to know what the cause is.  

Again, although I don't know the reason, there are also significant differences between individual tortoises in both carbon and nitrogen signatures (see below).  This might be because some are ranched and others wild-caught.  

To test this, I measured the degree of pyramiding in each tortoise as it is generally accepted that captive tortoises are more likely to show signs of pyramiding than wild tortoises (figure 3).  I then plotted signature over time of individuals with smooth, pyramided and intermediate carapaces, and statistically tested whether there was a difference in signature between the three groups (figure 2).  I found that there is a difference between smooth and pyramided tortoises in both carbon and nitrogen signature.

Figure 2: Changes in carbon and nitrogen signature over time. Each point is an individual with error bars showing 2 standard errors from the mean. Red: smooth carapace, blue: pyramided carapace, yellow: intermediate

Figure 3: smooth tortoise (top) and pyramided tortoise (bottom)

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).  

Figure 4: Carbon signature of over time. Each line is an individual. Red: smooth, Blue: pyramided, Yellow: intermediate. For this plot intermediate tortoises were allocated to the plot for group that they fell into in figure 2

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.