This is a slightly speculative post, and follows on from two previous articles which discussed the idea that red hair is the product of genetically diverse 'melting pot' cultures.
Those two articles can be found here;
In this post I'll add some further points. Particularly in regard the natural pigments responsible for hair and skin colour.
My general theory is that red hair occurs when people inherit high levels of pheomelanin from darker skinned/haired relatives (who have high levels of both eumelanin and pheomelanin), but low levels of eumelanin thanks to other lighter haired/skinned forebears.
i.e. that red hair is an imbalance that sometimes crops up when there is potential for very high and very low levels of these pigments in the gene pool
Eumelanin and Pheomelanin
Firstly, to give a brief overview I'll quote a few little passages from Wikipedia.
Eumelanin: There are two types of eumelanin, which are brown eumelanin and black eumelanin [..] A small amount of black eumelanin in the absence of other pigments causes grey hair. A small amount of brown eumelanin in the absence of other pigments causes yellow (blond) hair.
Pheomelanin: Pheomelanins impart a range of yellowish to reddish colors. Pheomelanins are particularly concentrated in the lips, nipples, glans of the penis, and vagina. When a small amount of brown eumelanin in hair, which would otherwise cause blond hair, is mixed with red pheomelanin, the result is orange hair, which is typically called "red" or "ginger" hair. Pheomelanin is also present in the skin, and redheads consequently often have a more pinkish hue to their skin as well.
Blond-haired people have low levels of both melanin types.
Again, Wikipedia; Blond hair can have almost any proportion of pheomelanin and eumelanin, but has only small amounts of both.
Another interesting fact I came across in my travels that may be worth remembering is that people with two variants of the famous MC1R gene are more likely to be blonde or fair-haired than ginger.
"MC1R only explains 73% of the SNP heritability for red hair in UK Biobank, and in fact most individuals with two MC1R variants have blonde or light brown hair."
So much for the much repeated "MC1R variant = redhead" then.
Bleaching The Hair
Something that adds to my view is what happens to hair when it's bleached. Apparently bleaching destroys the eumelanin more readily than it destroys pheomelanin, meaning the pheomelanin is left behind. Consequently dark hair is often left an orange colour after bleaching. Whereas fairer or blonder hair is left much whiter, as there's less pheomelanin to begin with.
The following article explains this point in a very easy to understand way.
If we assume that a dark hair strand has 10,000 "molecules" of total melanin, then by the ratios above, 9900 “molecules” (99%) of these would be eumelanin and 100 "molecules" (1%) would be pheomelanin. In comparison, a blond hair strand of equal size may have only 100 "molecules" of total melanin. Of these, 95 "molecules" are eumelanin (95%) and 5 "molecules" are pheomelanin (5%).
So though the blond hair has a higher percentage of pheomelanin, the dark hair still has more overall. Continuing;
When both dark and blond hair types are bleached, most of the accessible eumelanin in the two types of hair would be broken down and discolored, and only very little of the pheomelanin would do so. The reason why pheomelanin is harder to bleach is because it has sulfur linkages (or ties) that make the molecule more tightly packed and inaccessible to the bleach.The end result is that dark hair lifts to a brassy undertone because of the significant amount of pheomelanin left behind (let’s say 90 molecules), while blonde hair reaches the palest yellow stage because only very few pheomelanins were there to start with, and they get even less after the bleach (Let’s say 3 or 4 molecules).
I'm sure this is an oversimplification, as the article itself admits, but still, it nevertheles tallies well with my thoughts. Namely that redheads - through their genetics - have "naturally-bleached" dark hair in essence. They don't have the eumelanin, but are still left with a high amount of pheomelanin, which light-haired people wouldn't naturally have.
This Quora page asking "What does bleached black hair look like?" provides some handy illustrations. Likewise a quick Google Image search will bring up examples.
Freckles and Tanning
Another thing that perhaps links with this, though my thoughts here aren't as clearly delineated yet, is tanning and freckling.
It's not often noted, but freckling is in essence a type of tanning. Albeit sporadic. With freckles appearing on areas of the body exposed to the Sun. Such as the face, neck and forearms.
One difference with freckles though is that unlike normal tanning they appear to be permanent (though they may fade over many years). It makes me wonder if there are perhaps two types of tanning. A fixed permanent type, and a more flexible responsive type. Or perhaps simply two slightly difference types, loosely associated with the two types of melanin.
Both probably serve their purpose, and I would surmise that both are utilised by the human body, but that for people with a more standard balance between the two melanin types the tanning effects are much more seamless. Whereas for redheads the preponderance of one and near-complete lack of the other makes the visual impact of the first much more noticeable.
It would also perhaps make sense that permanent tanning would be more patchy (freckly) as with it being less reversible it would be overbearing if the entire skin changed colour permanently. Especially for people returning to a low sunlight climate after being exposed to the Sun.
Though it could also be the case that the small amount of pheomelanin in general means it can't go beyond freckly pinpricks of skin darkening. With the freckles of redheads being the body trying its hardest to produce a tanning response in the absence of the eumelanin which would normally do the largest share of the work.
I would guess that most people with darker hair/skin also have this capacity to "freckle tan", but that like with the hidden orange in the darker hair, only revealed by the bleaching, the effects are hugely overshadowed by the effects of the eumelanin.
So the pheomelanin is a minor supplementary factor in standard conditions. Seamless and blended in the palette of most, but starkly noticeable in redheads when it appears on its own. Or as more dominant than would be usual. In fact, the imbalance of pheomelanin to eumelanin relative to standard norms would also explain why you get darker and fairer redheads. Even redheads with dark skin. Such people have no doubt inherited high-end levels of pheomelanin, but more average or lesser levels of eumelanin. Leading to mid-range skin tone with freckles, or dark red hair, etc.
Whereas very fair redheads - who sometimes don't freckle at all - have relatively low levels of both pheomelanin and eumelanin, but still the pheomelanin is higher than average in relation to the eumelanin for what would be usual for people in that skin/hair tone band. Leading to bright, but light carroty red hair, or strawberry blonde perhaps. And fair skin with very little tanning at all - both as a consequence of eumelanin, or pheomelanin.
Domestication and Civilisation
Finally, one little point of speculation I feel is worth adding here is the role domestication or civilised living may play in hair variance.
In the animal kingdom we see that domesticated animals - cats, dogs, cattle, horses - often come in a multitude of fur/hair colours. In contrast to native species, where animals tend to look more fixed in type. It could be that the process of domestication - with humans largely deciding the breeding choices of the animals - means that sameness is less selected for.
Out in nature an animal born with an unusual colour pattern, or looking different in some other way, may be shunned by the wider group. Stopping its genes from passing on. Whereas in an environment where the selection pressures are different such unusual traits may be allowed to flourish - and things such as hair/fur colour may become less rigid and more liable to fluctuate.
I wonder if civilised living also has a similar effect on humans. In places where things are more dangerous and natural humans likewise may be killed or shunned by the wider group for looking different. Even today for example we sometimes hear tales of albinos being killed or persecuted in Africa for their appearance. However, when you're living somewhere with law and order, and a more peaceful way of living such persecutions becomes less frequent (though again, persecution for being different is not unknown, even in advanced places).
So perhaps peaceful civilisation allows variation in things like hair and eye colour to flourish and become less fixed and homogenous.
This is all speculation of course, but still, the melting pot idea seems quite enticing.