FaunaClassifieds - View Single Post - The odds gods, statistics, and patience with hets.
View Single Post
Old 07-22-2019, 10:29 PM   #1
nickolasanastasiou
The odds gods, statistics, and patience with hets.

As some people are aware, I have worked with some chelonian morphs for a decent number of years. One of these is the ivory (leucistic) morph of the sulcata tortoise.

In 2015, I added two juvenile/subadult het (heterozygous) for ivory females to my project from another party. I eventually transferred them to my project partner in TX since most of the land areas I have directly available to me in FL are not the best for Savannah species for one reason or another. Not terrible, but not ideal on these properties in particular. Perhaps a new property someday will change that.

For reference, we use only homozygous (visual) ivory sulcata males as our breeders. Neither male hets nor wild type (no ivory genetics involved) males are kept - much less provided access. There was no possible genetic contribution of anything other than Fife. bloodline ivory males.

Anyway, the females matured and then "seasoned" in 2016 and 2017, starting to produce bountiful fertile clutches as sulcatas tend to do. One of these females proved out in her first clutch by producing ivories as hoped. The other did not. The chance per egg, as an independent statistical event, of getting a homozygote (ivory) from ivory bred to het is 50%. The principle to understand here is that it is a coin flip each time and there is no result "owed" regardless of eggs hatching before or after any other egg. That does not mean a clutch of 20 should yield 10 ivories and 10 hets. Over a large enough sample size, it is assumed that things would even out, but my experience is that this is often not actually the case. I consider 40% of a clutch being ivory in this scenario to be pretty fortuitous and ~30% being ivory to be generally more realistic.

For the second female that failed to produce ivories, we were racking up the het baby results. We assumed after the first season that we could rule out the possibility of her being a het and we had planned to list her for sale this year as a possible het that was largely considered to be a non-het based on our results. She had given us 180 hatchlings, all hets, and then the improbable happened. An ivory emerged. We were so surprised and it was sort of extra pale in appearance compared to many of our ivories, so we disbelieved it at first or thought it might even be something spontaneously new. We track and separate all of our eggs by clutch. No mixing between females occurs when we collect eggs and incubate. There were no mistakes in identifying the source. The sires had not changed. The dams had not changed. In subsequent eggs, this female has since thrown us a few more ivories. Not nearly as many as her sister, but infrequent sprinklings here and there.

I share this firstly because it is cool. The statistical chance as a string of 50% chances 180 hatchlings strong is impressively improbable, yet it happened. I used to use a metric of 20 consecutive results in order to functionally rule out a supposed het's genuine het status unless I produced it myself (which would allow me absolute confidence in the ideal scenario). I now have to revise that metric to two full seasons of results before I can make a call in good faith on a "disproven" het (although one can never truly disprove a het by math; the chance simply infinitely approaches zero without reaching it). I share this secondly because I see people quick to jump the gun time and time again with snakes, lizards, turtles, and so on regarding hets. Do people sell and buy fake hets? Often. VERY often, in fact, and it is quite a shame. However, I see people thinking they deserve an early win going into a game of statistical chance and then taking a less fortuitous set of results with a relatively small sample size to the BOI or other settings to gripe. Your three eggs did not yield an albino with hets involved? Your seven eggs did not yield a piebald with hets involved? Your ten or twenty eggs did not yield a double visual with hets involved? These scenarios have disproven nothing regarding the legitimacy of the het label assigned to a given parent. You may have a non-het that was sold to you under false pretenses. You may also have a perfectly good het and perfectly bad luck. One's bad luck is not a justifiable reason to bash someone's reputation, so one should keep perspective on this. If you want only good luck, buy all visuals. If you want statistical chance with acceptance that not every egg will go your way while gaining affordable access to a morph, dabble in hets.

I have never heard of anyone having to wait for their 181st result to prove out a morph before. That is why I share this. It is an interesting (and uncommon) case and something to take into account when exercising patience and proving out a het.

One of the sires, two dams (including the one in question), the 181st baby, and some later babies are shown.
Attached Images