1. A curious enigma
This is one of those facts of biology which provides prime material for comedians to make an ‘easy punch below the belt’ (pun intended) against the notion of ‘intelligent design’.
Was the human body designed by committee? If you play with the balls, the penis likes that. It’s kind of fun. But we did find out one negative thing about the balls. If you hit them really hard… it’s a total system reset. It’s like… If it was a slot machine, it would pay. It was kind of rough.
Robin Willams (see full video clip)
Jokes aside, our main question is still intriguing, and finding a good answer is more difficult than one might expect. Why bother putting testes in a scrotum only to expose these sensitive gonads (and you chances of reproducing) to the dangerous world? This is also one of many things that makes mammals oddballs among animals. In all vertebrates (including mammals), the gonads develop deep within the body right next to the kidneys, and in most vertebrate species the gonads remain (more or less) in place throughout life. In fact, the urinary and reproductive systems are very closely linked (by location and development), which is why anatomists often group both together under the ‘urogenital system’. Yet another biological fact exploited for humor.
“I think of, like, the human body, and I look at what’s going on between our legs. Down there between our legs, it’s like an entertainment complex in the middle of a sewage system”
Neil deGrasse Tyson
But why not just keep the gonads there? The relocation of the testis during development also leads to weak spots in the abdomen (the inguinal canal), which can result in painful inguinal hernias. The descend of the testes can also malfunction, such that some are born with undescended testicles which can persist throughout life leading to infertility. But, even if they have ‘faithfully’ descended, the testes end up being suspended on strings, making them vulnerable to ‘spontaneous’ injuries (torsions). No kick in the nuts required. Most would suspect that there must be some benefit which justifies these tradeoffs.
The common answer I see is the following: Mammals are warm blooded, but testes require low temperatures to produce and store healthy sperm. So, the scrotum is a solution to relocate the testicles away from the hot interior of the body. The scrotum is also controlled by smooth muscles, moving the testicles closer to or further away from the abdomen, regulating the temperature of the testes. This is called the ‘Cool Spermatogenesis’ hypothesis, which was first proposed by CR Moore in 1926. There are facts that are commonly cited to support this. Sperm cells are indeed very temperature sensitive. There are a few studies which found that human sperm count were different between seasons, though the magnitude of the difference may not matter much in practice (here is a recent meta analysis).
This seems reasonable at first, until you start to notice the many glaring exceptions. (1) Not all mammals have a scrotum. Although, one could be quick and point out that these exceptions actually support the rule of he ‘Cool Spermatogenesis’ hypothesis. Mammals that lack a scrotum tend to have lower body temperatures, such as tenrecs. However, this “low body temperature means no scrotum needed” principle is far from consistent. (2) There are mammals with rather high body temperatures and internal testicles. (3) It also doesn’t explain why no bird species have externalized their testes in some manner, despite the fact that they are generally hotter than mammals.
So, we can broaden the question even further: Why do only (some) mammals (with high body temperatures) have a scrotum, and why are there no birds with a scrotum(-like organ)? Are there any alternative or complementary hypotheses that can explain this conundrum?
2. Who has a scrotum?
Before we answer this, we also need to establish that there are various degrees of descended testes. They can be ‘completely’ descended into a scrotum, or they can be ‘partially’ descended (without a scrotum), or they are ‘undescended’. There are several papers that discuss this in detail, but this paper provides the following image summarizing the distribution of descended testis among mammals.
CDT = “completely descended testes (scrotum)”
IDT = “incompletely descended testes (no scrotum)”
UDA = “undescended testes (no scrotum)”
Everyone agrees on the distribution depicted here, but there are disagreements on when scrota (or scrotums) first appeared in evolutionary history. To fully unpack this, we also need to cover some phylogenetics. First, mammals split into two groups:
- Monotremes: Only 5 surving species: 1 Platypus and 4 Echidnas.
- Therians: Which in turn splits into the following two groups
- Marsupialia: e.g. kangaroos, wallabies, possums, koalas, wombats, bandicoots, Tasmanian devils, etc
- Placentalia: Includes ~90% of all mammal species, which in turn splits into…
- Xenarthra: sloths, anteaters, and armadillos.
- Afrotheria: elephants, manatees, hyraxes, aardvarks, tenrecs, elephant shrews, and golden moles.
- Boreoeutheria: ~97% of all placental mammals: rodents, bats, whales, horses, cows, sheep, dogs, cats, mice, you (and many… many others).
The monophyly of each of these groups is very firmly established. However, how the last three placental mammal groups are related to each other is somewhat contested. Different genetic studies recover different relationships. It appears that these three lineages diverged from a common ancestor within a relatively short time frame (as in within 10 to a few 10s of millions of years). This can lead to incomplete lineage sorting, which causes conflicting results in phylogenetics between different datasets. Nevertheless, most recent studies gives more support the ‘Atlantogenata’ hypothesis, with Afrotheria and Xenarthra closer to each other than either are to Boreoeutheria.
- Placentalia
- Boreoeutheria:
- Atlantogenata:
- Xenarthra
- Afrotheria
This relationshp is also shown in the image above. Apologies for the nitty gritty details, but this is relevant to our discussion. While Xenarthra and Afrotheria are each characterized by a set morphological traits, the members within Boreoeutheria have almost no uniquely shared characteristics. To my knowledge, the only common anatomical trait is the scrotum, while Afrotheria and Xenarthra have either only partially descended or undescended testes. In fact, one group within Boreoeutheria is called ‘scrotifera’. Still, scrota are far from a universal trait among Boreoeutherians. There are quite a few exceptions (listed in the next paragraph). But there some non-anatomical commonalities that do apply. One of these is their paleo-biogeography. Boreoeutherians originally lived on the northern super continent Laurasia; the super continent that existed when North America and Eurasia were a combined landmass. Hence the name ‘boreo-eu-theria’, which translates to ‘northern-true-beasts’. Of course, many Boreoeutherians have migrated since then and are now found on every continent. On the other hand, Xenarthra and Afrotheria originated on the Southern continents; South America and Africa respectively. The combined ‘Atlantogenata’ group is named after the fact that the two ancestral continents are separated by the Atantic ocean. Another peculiar detail to note is the fact that (as far as I am aware) boreoeutherias are the only mammals susceptible to lentiviruses, which is a genus of viruses that includes HIV. Lucky us 
.
As previously mentioned, most Boreoeutherians do have a scrotum with the following exceptions with incompletely descended testicles (source and source): Some rodents (members of Hystricomorpha and Spalacidae), some Chiropterans (bats), all Ceratomorphs (tapirs+rhinos), all Pangolins, all Whippomorphs (hippos+whales), all Eulipotyphlans (Shrews, moles and hedgehogs), all Phocids (earless seals) and all Odobenids (one species of walrus remaining), BUT curiously the Otariids (eared seals) do have scrota. Since Odobenids and Otariids are more closely related to each other than either is to the Phocids, that likely means Phocids and Odobenids lost scrota independently.
Lastly, we have two remaining lineages of mammals: the Marsupials and Monotremes. Marsupials are more closely related to the placentals, and it turns out that almost all of them do have scrota. The only exception is the marsupial mole (see the recent paper on them). On the other hand, no Monotreme has a scrotum, nor do they have partially descended testes.
3. So, when did scrota evolve?
This is a difficult question since (1) as noted the presence of scrota isn’t unique to one group with all members possessing it. The scrotum must have evolved once (maybe even multiple times) and have been lost multiple times. Thus, there many different scenarios that can explain the distribution of this trait; and (2) scrota and testes do not preserve in the fossil record, so we can’t figure it’s evolution using that. Still, most research that I have seen conclude that scrota with completely descended testes is the ancestral condition for Boreoeutheria, with many independent losses within the group.
Next, we have Atlantogenata (Afrotheria + Xenarthra). None of these have a scrotum. So one conclusion we could make is that their ancestors never possessed scrota. However, we also see that some members do have partially descended testicles. So, is this trait plesiomorphic (ancestral)? i.e. did the ancestors of all placental mammals have partially descended testes, which was retained in Atlantogenata while Boreoeutheria evolved completely descended testes? Or, is this trait apomorphic (derived) to Atlentaogeata? i.e. he ancestors did not have partially descended testes, they either have descended or completely descended, while Atlantogenata evolved partially descended testes.
An answer might come from the next group: the Marsupials. As mentioned, almost all of them have scrota. So, if the scrotum evolved only once, that must mean that it evolved in the ancestors of all Therians (marsupials+placentals). This means that the scrotum is a plesiomorphic trait of placentals, and was later lost in Atlantogenata, and of course among various Boreoeutherians. However, not all researchers agree with this scenario. Some suspect that the scrotum of marsupials was independent from that of Boreoeutherians for a few reasons. For one, there is a big anatomical difference. The marsupial scrotum is placed in-front of the penis, whereas (as you may know) in placentals it’s placed behind the penis. Furthermore the molecular signaling that initiate the migration of the testes is also different.
(image source)
The journey of the testes into the scrotum occurs in two phases. The first phase involves the descend of the testis within the abdomen: The Transabdominal phase. The descend is mediated by the expansion of the gubernaculum, a ligament that is attached to the bottom of the gonads. This is also true for the ovaries to a degree, which also moves a significant distance from their original spot. Regarding the testes, this is followed by the testes passing through the inguinal canal into the scrotum: The Inguinoscrotal phase.
In placentals, latter Inguinoscrotal phase is induced by hormones, androgens in particular. But the development of the scrotum in Marsupials is dependent on x-chromosome dosage. This means that intersex XXY (Klinefelter Syndrome) placentals mostly develop a scrotum, but often one or both testes have not moved into the scrotum at birth. In contrast, XXY marsupials tend to have a pouch, phallus, but no scrotum. Furthermore, intersex XO (Turner syndrome) placentals don’t develop a scrotum, but intersex XO marsupials do have a scrotum while possessing ovaries, but no pouch (source).
Nevertheless, there are distinctive similarities between marsupials and placentals regarding the former transabdominal phase. INSL3 (insulin-like 3) is the key hormone produced by Leydig cells of the testis. INSL3 binds to the RXFP2 receptor to initiates the movement of the testes. Marsupials also have the two genes encoding these hormone and receptor proteins. And it turns out, INSL3 and RXFP2 became pseudogenes multiple times within Afrotheria, and their testes do not descend at all. The aardvark is the only Afrotherian with partially descended testes. Then we are left with the Monotremes, which do not have desended testes, and they also don’t have the gubernaculum nor hormone-receptor interaction between INSL3 and RXFP2.
So, putting this all together, this could indicate that incompletely descended testes was likely present in the ancestors of all Therians, but not the monotremes. Latter, partially descended testes was later lost many times among Afrotherians. Furthermore, unless we are dealing with developmental system drift, the differences in anatomy and the physiological triggers for scrotum development suggests that Marsupials and Boreoutherians independently evolved scrota (with multiple losses within the groups). Again, I should stress that this is just one of many hypotheses. For example, the paper (which is the source of that phylogenetic figure I added above) argues that scrota did evolved only once in the ancestor to all mammals, even the monotremes. Personally, I find the former hypothesis more plausible, but that’s just my intuition.
Nullius in verba.
4. Okay, but WHY though?
Previously, I mentioned the familiar ‘Cool Spermatogenesis’ Hypothesis (CR Moore in 1926) which argues that cooler temperatures are required for making healthy sperm… but why? There are other (non-mutually exclusive) hypotheses that addresses this question. You have the ‘Mutation’ Hypothesis (RV Short 1997), proposing that cooler temperature decreases germline mutation rates. Besides this we also have the ‘Activation’ hypothesis (Gallup et al. 2009), which says that by keeping the sperm cool, the increase in temperature during intercourse can function as an activating signal. There is also the ‘training hypothesis’ (Freeman 1990), which does not cite temperature. Instead, it argues that the scrotum is a way to decrease blood supply to the testis, such that the sperm are ‘trained’ to cope with low oxygen levels, like how muscle cells respond to lack of oxygen by enhancing aerobic metabolism (via increasing the number and size of mitochondria). If sperm adapt in a similar manner, perhaps this increases their chances of surviving the journey to the egg cell.
However, none of these hypotheses account for the fact that many hot mammals don’t have scrota, nor why no bird have externalized their testes. It could also be the case that the testes were externalized for a different reason, and testes simply re-adjusted to function at cooler temperatures afterwards. Or even if cooler temperatures are needed, these hypotheses do not explain why the testes cannot be cooled (or supplied with less blood) while still tucked safely inside. According to the ‘Epididymis cooling’ hypothesis’ Bedford 2004, the point of cooler temperature only applies to the epididymis; the tube near the testes where the sperm cells are matured and stored. To support this, Bedford points out that (1) many scrota that have a lot of fur on them still have a tiny bald patch directly above the epididymis; and (2) many mammals lacking a scrotum have positioned their epididymis just below the skin to cool them down.
There are a few other hypotheses though. Two (or one) of them are the ‘Display’ and ‘Handicap’ Hypothesis (Portmann 1952). These cite Sexual selection as the driving mechanism for the evolution of the scrotum; which evolved either as a means of sexual display to attract mates, and/or as a handicap to function as an honest signal of fitness to the opposite sex. There are a few species with pigmented scrota, most famously the blue pigmented scrotum of the vervet monkey. However, this hypothesis isn’t widely accepted since such extravagant scrota are not common. Most are rather inconspicuous. There is no disagreement that the brightly colored scrotum of mammals like the vervet monkey is a sexual display, but that’s probably not why the scrotum itself evolved. The scrotum has been later co-opted to function as a canvas by some mammals.
The last (and in my opinion the best) explanation to discuss is the ‘Galloping’ Hypothesis (Frey 1991 and MRA Chance 1996). This argues that the scrotum is a way for the testes to avoid the hydro-static pressure fluctuations within the abdomen due to galloping (or jumping, or leaping, etc). Chance proposed this explanation after he read a newspaper story about the Oxford-Cambridge University boat race. Apparently, some researchers tested the urine of these oarsmen students, and they found that their urine contained fluid from their prostates. Why is this important? Well, there are in fact no sphincters in the reproductive tract nor any other means to act as a ‘valve’. Without this, fluids can freely move due to squeezing the tubes and sacs of this tract. The argument is that if the testes remained inside, sperm would leak out (or the testes can get damaged) from pressure changed whenever the mammal started to gallop.
Frey provides some evidence in support of this. He observed features of blood vessels of scrotal testes that ensure more constant pressure, possibly to avoid impaired blood drainage during galloping. He also noted that the specific adaptations are different between marsupials, further indicating that Marsupials likely evolved the scrotum independently. Furthermore, if we list the mammals that gallop/run/leap they almost always have scrota, and vise versa. For example, elephants (lacking scrota) don’t gallop or jump; and when they ‘run’, they don’t really run. It’s more like a very fast shuffle. Burrowing mammals have also internalized testes, most notably the Marsupial mole, the only marsupial without a scrotum.
And aquatic mammals also lack scrota, with the curious exception being the Otariids (eared seals), while all other seals are ascrotal. However, Otariids are notably different from the other Pinnepeds regarding how they move. While earless seals and the walrus use their hind flippers for propulsion, and their front flippers for steering; Otariids flap their front flippers to propel themselves, and flex the rest of their body to be very agile in the water. To make sudden turns, some can flex their spine such that their head almost touches the hind-flippers. Otariids are also the most capable to ‘run’, and you can look up videos of them chasing after penguins ON LAND.
What about Rhinos and Tapirs? They can gallop, but they don’t have scrota. However, their testes are still outside the abdomen; located in a subcutaneous scrotal-like sac! The only mammals that break the gallop rule are the elephant shrews (Macroscelidea). These are Afrotherians, none of which have scrota, but elephant shrews are still very fast runners. Among the fastest of all small mammals. Perhaps they are too small (largest is still just 700 grams) for abdominal pressure fluctuations to really matter? And what about birds? They do not break the rule. Fast moving mammals move very differently than fast moving birds; even ground runners like the ostrich and roadrunners. Mammals flex their spines up and down, while the thorax of birds are rather stiff. To me, the ‘galloping’ hypothesis seems the most plausible explanation. But, again, that’s just me.