New thoughts on the fossil faunas of Mediterranean islands

(writing in progress)

Gymnesic islands of Balearics uniquely small and isolated among world's islands inhabited by endemic bovids
 
Rozzi & Palombo (2014), on insularised bovids, seems to miss a point. This is that all of the other islands inhabited by (extremely limited faunas of) bovids are either large islands or dubiously isolated from the mainland. Please see the map below to make the point that Majorca and Minorca, even when joined at low sea levels, are much smaller than any of the other islands supporting endemic genera/species of bovid.
 
Japan has an endemic species of serow but consists of large islands, which formerly were large enough to support (a small form of) the wolf Canis lupus. Taiwan is smaller, but is a continental shelf island joined to the mainland during the glacials, and anyway is still larger than the Gymnesics (Majorca and Minorca). Far from being similar to the Gymnesics in size, Sulawesi (supporting the endemic genus Anoa) is more comparable to the whole of peninsular Italy. Sulawesi remains a relatively large island even considering its peninsularisation into several quasi-insular units. And the fact that macaques and tarsiers reached Sulawesi casts some doubt on the degree of isolation from Borneo, which is part of the continental shelf and thus part of mainland Asia for much of the Pleistocene.
 
In the case of Bubalus mindorensis, restricted to Mindoro island in the Philippines: in the first place this species is not much different from mainland congeners, being only slightly dwarfed with a body mass of up to 300kg (i.e. more massive than any southern African bovid other than buffalo, eland and roan antelope). In the second place Mindoro, although similar to the Gymnesics in area, is a continental shelf island in the sense that it would have been connected to other Philippine islands at low sea levels during the glacials. And the Philippines are even less clearly isolated from mainland Asia than Sulawesi is, because Palawan provides a connection. So B. mindorensis is ambivalent in both ways, i.e. a rather poor example of insular evolution of bovids.
 
My point: the Gymnesics (Majorca and Minorca), even at the glacial low sea levels, are unique small among islands inhabited by endemic bovids.
 
Even if the ancestor of Myotragus (which should  not really be called a sheep because it more resembled a goral or serow, which is a rather nondescript antelope-like bovid albeit related to sheep and goats) had reached the Balearic islands only within the Pleistocene, the small size of these islands and the extreme isolation (with no prospect of genetic exchange with the mainland) would lead us to expect drastic adaptive evolution or else extinction. The extreme modification of Myotragus balearicus would make sense even if the genus was only one million years old. Instead the genus is more like 5 million years old and it’s only really the most recent species, M. balearicus, that is drastically modified.
 
Why did it take Myotragus so long (about 5 million years) to become drastically modified to what, in the scheme of things, is a drastically unusual and testing environment for any bovid? This slowness to adapt would seem to be more interesting than the fact that it has at last adapted. If such adaptations have proven necessary for the survival of a bovid on such a small and isolated island, how did the less specialised, earlier species of Myotragus survive at all?
 
To me there are only two possible answers: either evolution really is constrained in its rate, which is basically irrational because it implies that organisms can survive for long periods in maladaptation to their environment, or what has happened is that the physical enviroment of the Balearics has changed for the ‘worse’ during the last, say, million years. Myotragus was perhaps able to survive for something like four million years with only modest adaptation because these islands were not all that hard to live on. Then the climate changed and it became so hard to survive that M. balearicus arose, with its unprecedented features.
 
So whereas many may assume that M. balearicus is the way it is owing to insularisation, what I’m suggesting is that it’s really the way it is because of the unique nature of its islands over the last million years: both extremely isolated and extremely poor as an environment for bovids (in total land area and the climate during the second half of the Pleistocene).  
 
The following map shows just how small the Gymnesics are compared to the other islands in the world where endemic genera/species of bovids occur/occurred:
 
http://thumbnails-visually.netdna-ssl.com/worldmap-worldmap-photos-wallpapers-galleries-full-hd_50290fb555fd4_w1500.jpg
 
Mindoro, the Philippine island inhabited by the endemic bovid Bubalus mindorensis:
https://en.wikipedia.org/wiki/Mindoro#/media/File:Ph_locator_map_mindoro.png

Comparison of gymnesics with Cyprus, as original interpretation of Mediterranean insular faunas
 
The following is an angle that nobody seems previously to have come up with, on the topic of the recently-extinct insularised faunas of Mediterranean islands.
 
The most isolated of the Mediterranean islands are the most interesting, because most of the islands in the Mediterranean Sea have been intermittently connected to the mainland for much of the last two million years. The ambivalent isolation of most Mediterranean islands makes them unsatisfactory ‘natural laboratories’ in which it is unclear just what is ‘treatment’ and what is ‘control’. It is the oceanic, as opposed to continental, islands in the Mediterranean which offer the greatest biological insights because they have provided a clear natural experiment in isolation and adaptation.
 
The Gymnesic islands (Majorca and Minorca, which have been joined for much of the last two million years during low sea levels during Pleistocene glacial periods) lie in the far west of the Mediterranean Sea, whereas Cyprus lies in the far east of the Mediterranean Sea. These are oceanic islands because, even at the lowest sea levels of the Pleistocene, they remained isolated from the mainland to a degree that Sicily, Sardinia, Corsica, Crete, Malta, and the hundreds of small Greek islands did not. The Gymnesics and Cyprus have similar land areas, making them even more comparable.
 
When we put the Gymnesics and Cyprus together as a logical comparison, something emerges which I’ve not seen pointed out clearly before: their faunas could hardly be more different.
 
The Gymnesics were inhabited by a small bovid whereas Cyprus was inhabited by miniaturised hippo and elephant. Furthermore, the Gymnesics were inhabited for 5.3 million years, during which a whole series of chronospecies of Myotragus evolved in-situ, whereas Cyprus was inhabited for no more than 0.25 million years, lying empty of herbivorous mammals for most of its history; and in Cyprus the process of adaptation to island conditions seems to have been virtually instantaneous.
 
Both lineages, the bovid lineage on the Gymnesics and the hippo and elephant lineages on Cyprus, were abruptly exterminated when humans settled on these islands in the Holocene. This is a turn of events typical for oceanic islands worldwide, and is the least puzzling aspect of this topic.
 
Much has been written about the original faunas of both islands, but a certain cognitive dissonance remains that I’m the first to acknowledge, here.
 
There seems no accounting for the faunistic differences except pure chance, which is theoretically unsatisfactory.
 
Everyone accepts that the Gymnesics were populated by the ancestor of Myotragus by means of the improbable but credible land connection during the Messinian salinity crisis, about 5.5 million years ago. But nobody has explained why a similar population did not occur on Cyprus at the same time. The best attempts to explain the failure of ungulates to colonise Cyprus during the Messinian salinity crisis invoke a particularly harsh salt-desert in the eastern Mediterranean, making proto-Cyprus an island even when the Mediterranean was largely dry, surrounded by a dry desert as opposed to a sea barrier. This explanation seems arbitrary and convenient, would you not agree?
 
Everyone also accepts that Cyprus was (eventually, about 5 million years later!) populated by the ancestors of its hippo and elephant by transoceanic dispersal, i.e. swimming. But nobody has explained either why it took so long for hippo and elephant to swim to Cyprus, or why hippo and elephant did not likewise swim to the Gymnesics, usurping the niche occupied by Myotragus.
 
So the accepted faunistic histories of these islands remain ‘just-so stories’ lacking in any plausible rationale of cause and effect, not so?
 
There is a rationale of cause and effect w.r.t. the eventual drastic adaptation of these animals to the isolation of their island habitats. Myotragus on the Gymnesics is perhaps, overall, the most oddly adapted bovid on Earth (small and forward-facing eyes, just one pair of incisors, incisors ever-growing, bone-growth interrupted, lifespan extremely long, foot-bones extremely short). Phanourios minor (the Cyprus hippo) and Elephas cypriotes (the Cyprus elephant) are both far smaller than their closest relatives on nearby mainlands have ever been, showing great diminution in body size as the main adaptation to island life. In this sense, isolation did affected the ungulates on both the Gymnesics and Cyprus in ways that make sense ecologically and evolutionarily. But where cause and effect are lacking is in why such different lineages came to populate these islands, and how Cyprus could have lain for 5 million long years without any apparent mammalian herbivore, a perplexing case of an ‘empty niche’.
 
Furthermore, there is some cognitive dissonance in believing that it took 5 million years for the Myotragus lineage to reach its full degree of adaptation to island conditions, whereas this process of appropriate adaptation was completed by the corresponding fauna of Cyprus virtually overnight in evolutionary terms. Why did Myotragus balearicus, the most recent and most drastically modified form, not evolve already 5.2 million years ago.
 
So I’ve pointed out two ‘lags’ in this scenario, neither of which can be accepted as mere accidents if we’re to understand cause and effect in biology. First there is the lag in any ungulate reaching Cyprus at all, a lag that lasted 5 million years and meant that the resources of Cyprus were left unused for a period far exceeding the whole of human evolution from ape to modern scientist. Secondly there is a lag in the full adaptation to its habitat of the ungulate found on the Gymnesics.
 
The bottom line is that neither the biological nature of the bovids, hippos and elephants, nor the geological/geographical history of the Mediterranean, can explain why such different animals came to live on the Gymnesics and Cyprus. Here we have a biogeographical and ecological mystery which deserves to be acknowledged.
 
Additional herbivores present in western Mediterranean islands
 
According to this book https://books.google.com.au/books?id=nkLjBwAAQBAJ&pg=PA175&lpg=PA175&dq=time+of+human+arrival+on+mediterranean+islands&source=bl&ots=B79OXksa9r&sig=FE5huEw6W0a7XQIbd_9RWZDDlsU&hl=en&sa=X&ved=0CBwQ6AEwAGoVChMI366amIjVyAIVQb6UCh1mJAjZ#v=onepage&q=time%20of%20human%20arrival%20on%20mediterranean%20islands&f=false
the following additional herbivorous vertebrates occurred on islands in the western Mediterranean.
 
This book is also a reference to the fact that Myotragus balearicus went extinct about 4300 years ago, after perhaps 4000 years of coexistence with humans.
 
In the Pityusics of the Balearics (namely Eivissa = Ibiza), there was a caprine bovid (something similar to a goral) and a giant tortoise in the Pliocene, after the Zanclean Flood. Unlike the situation in the Gymnesics, this caprine died out. What replaced the bovid and the giant tortoise in the Pityusics was a species of Anser called the Pityusic goose. This goose, which may be an extinct species, seems on the basis of fossil material to have been abundant on Eivissa = Ibiza. Geese (Anser) were of course also present in the Gymnesics (after all, they can simply fly in every year) but the evidence is that their niche was largely precluded by Myotragus there.
 
Minorca, in the Gymnesics, had a giant tortoise plus a lagomorph before the niches were usurped by Myotragus.
 
On Sardinia and Corsica at the end of the Pleistocene and the beginning of the Holocene, a major herbivore seems to have been Megaloceros cazioti, a relative of the ‘giant Irish elk’. This island species was about impala-size, in keeping with the general principle of diminution of body size on islands, but it was otherwise typical of the mainland fauna which makes sense because a) these islands are relatively large, b) there were intermittent connections to the mainland in the glacials, and c) humans were probably present on these islands for tens of thousands of years before the Holocene, and the reproductive rates of these deer would have allowed insular coexistence between humans as predators and deer as prey.
 
Two realisations emerge, for me.
 
Firstly, both giant tortoises and geese (Anser) are good ‘default’ herbivores on Mediterranean islands because both are so dispersible in their different ways, and so robust in their ability to survive island conditions. Tortoises survive by sheer tolerance, having minimal demands. Geese survive mainly by not being confined to the islands, although of course it’s possible that the Pityusic goose was endemic and restricted to the Pityusics.
 
Although there seems to be no evidence of giant tortoises on Cyprus, the inference – which I’ve not seen anyone make previously – is that for the five million years before hippo and elephant arrived on Cyprus it’s perfectly possible that the main herbivores there were geese (whether these left a fossil record or not).

Why no wild pigs gone insularised?
  
Here’s something hiding in plain sight on the topic of the Pleistocene fauna of the Mediterranean islands.
 
The fact that diminutive hippos and elephants occurred on Mediterranean islands is explained by the swimming ability of these mammals. This is held up as an implicit explanation of the success of such incongruous, specialised animals in what are often dry and rocky habitats on the islands, and the evolutionary pressure on them to reduce their body size rapidly.
 
But on reflection this doesn’t make much sense. All along there has been a more suitable candidate with comparable powers of swimming: the wild boar Sus. Given that the wild boar is not far behind elephants (and even hippos) in swimming ability, and occurred on the Mediterranean mainland in the Pleistocene as elsewhere in Eurasia and North Africa, why on Earth did it not colonise at least the closer islands such as Sardinia and Crete, evolving on those islands into a modified form more suited to herbivory than to its original omnivory. Surely this is more plausible than survival of hippos in such situations? After all, wild pigs are already of suitable size to survive on islands, have broad diets which allow them to survive shortages of food, and can easily be imagined as evolved far enough to resemble a pygmy hippo.

Prof. Mb: “Majorca and Minorca are the only islands, worldwide, on which wild relatives of goats and sheep – which are poor swimmers - have survived[Footnote: resulting in the eventual evolution of the extremely island-adapted Myotragus balearicus, hardly recognisable as a relative of sheep or goats] for five million years in complete isolation from mainlands. The simple reason: the original colonisation occurred more than five million years ago, when the Mediterranean Sea dried up[Footnote: This Messinian Salinity Crisis ended 5.3 million years ago when the Mediterranean Sea once again filled with water.] so much that it was possible to walk to these islands.” (LINEAR LOGIC)
 
RHB: “Although there is little doubt that the ancestors of Myotragus walked to Majorca and Minorca, this does not explain their persistence there until exterminated by humans over a period so long that it exceeds the whole of human evolution from apes. The question remains of why hippos and elephants, which colonised many other Mediterranean islands by swimming, did not usurp the niche of herbivore after reaching Majorca and Minorca.” (LATERAL THINKING) 

Re-thinking the biogeography of mediterranean islands from scratch on sound conceptual basis:
 
Essentially the more we think about ‘what could be’ on the Mediterranean islands, the better. Instead, this seems to have been the last question on the mind of any of our peers.
 
In the Mediterranean Sea there are many islands of various sizes and various distances from the present mainland. All of them are, to some degree, natural laboratories that can teach us about biological and ecological principles. But the whole literature on this topic seems mired in a combination of complexity and woolly thinking. This is why it would pay us to put aside everything interpretive that’s been written before, marshal the facts as best we can, and think the whole thing through for ourselves.
 
Because most Mediterranean islands are ambivalent as islands (not being fully separated from the mainland at various times in the past), it’s wise to focus on the most isolated islands in the Mediterranean sea, viz the Gymnesics (Majorca and Minorca, which have been one island for much of the Pleistocene) and Cyprus (which has been separated from the mainland by about 100 km, a distance similar to the sea barrier between Australia and Asia (i.e. the Timor straits during lowest sea level). I emphasise this point: although Cyprus seems to be nestled in a corner of the Mediterranean, almost snuggled among the sheets of Asia Minor, it’s actually as remote from the mainland as Australia has been from Asia – with the full weight that this has carried in people’s minds in terms of the separation of major biotic realms of the world. Cyprus is effectively an oceanic island despite being only one of many Mediterranean islands, and despite not being located in the middle of the Mediterranean Sea.
 
Once we understand the Gymnesics and Cyprus, it will be easier to understand all the other islands.
 
Now, if we start from a rational perspective, knowing the mainland fauna of Europe and North Africa and the swimming abilities of the various mammals, it would not be difficult to predict which lineages were most likely to reach Mediterranean islands across sea barriers of various widths up to and including the approx. 100 km between Asia Minor and Cyprus. These are, surely, pigs (Sus), water buffaloes (Bubalus), geese (Anser) and tortoises.
 
To my mind there’s nothing difficult about this prediction. We know that the wild boar (Sus scrofa) swims well, because we’ve caught it about 10 km out to sea in both the Mediterranean and the English Channel. We know that water buffaloes swim well, because in southeast Asia they take the place of hippos in their habit of resting in the water and grazing near to the water. We also know that water buffaloes have crossed Wallace’s Line to get to Sulawesi from Borneo, and crossed shallower but comparably wide sea straits to get to the Philippines from Borneo. Water buffaloes no longer live wild in Europe today, but they were widespread here (in the form of Bubalus murrensis) until the Holocene.
 
As for geese, Anser is completely in line with Sus and Bubalus in being a widespread, common herbivore domesticated by humans, and completely indigenous to the Mediterranean Basin on a seasonal basis. It has the advantage of being able to fly to any island in the Mediterranean each year, and then leave again each year if necessary. On the other hand, it might want to stay in view of the freedom from predation that any insularised herbivore can enjoy on islands, which are predictably predator-free to a large degree.
 
We also know from elsewhere on Earth that herbivorous tortoises reach islands relatively easily by floating and swimming, and tend to survive on islands, often in gigantic form.  So I would predict relatively large terrestrial tortoises on at least some of the Mediterranean islands.
 
I’d be confident in predicting that a major feature of the large non-reptilian fauna of most if not all Mediterranean islands would be descendents from Sus, Bubalus and Anser. I’d not be surprised to find these animals evolved into different species or even genera, in adaptation to the insular conditions. For example I’d not be surprised to find Mediterranean island pigs evolved into more strictly herbivorous forms than S. scrofa, and named ‘Novosus’ or some such. And I’d not be surprised to find Mediterranean island water buffaloes in miniature form, suited to the limited resources of islands. But I’d have no trouble predicting the success of the descendants of these genera because not only can they swim to reach islands but their body form (unspecialised locomotorily) suits them as well to climbing rough slopes as to swimming long distances. I’d have no trouble envisaging an island-endemic pig suited to grazing on the rocky slopes of Crete, or a diminutive water buffalo covering the full range of landforms on Corsica. And I’d not be surprised to find a flightless goose on some of these islands, and if not then just the common old greylag (Anser anser) visiting each year and going to breed in northern Europe as usual.
 
Instead, I’d be surprised if such animals did not exist on these islands.
 
What else would I predict on these islands?
 
Well, we know from the general biogeography of deer (Cervidae) that they seem to swim well to islands, and do well on islands once they reach them. There are various examples of deer of reduced body size in southeast Asia, for example, including islands as remote as the Philippines. So it would be fair to expect deer, in adapted form, on at least some Mediterranean islands, not so? But I’d like to point out that, as far as I know, no Eurasian lineage of deer can match water buffaloes (Bubalus) in degree of adaptation to water.
 
Having made clear predictions, let’s turn to reality.
 
The reality of the faunas of the Gymnesics and Cyprus is surprising in several ways. In fact, almost everything about the herbivorous mammalian faunas of these islands is surprising to me.
 
On the Gymnesics, the real fauna consisted of bovids of a lineage with a poor record, elsewhere, of transoceanic dispersal: the caprines. Caprines are well-adapted to broken terrain but are regarded as poor swimmers. I can’t think of a single island, anywhere on Earth, that has been naturally colonised by caprine bovids across a sea barrier. The body form of caprines should not, in principle, prevent them from swimming as well as most ruminants do, but there seems to be something behavioural hardwired into them as part of their overall specialisation on mountainous terrain. They have minimal attraction to water and thus differ from bovine bovids such as water buffaloes.
 
It may confuse the lay person that, if one travels to Cyprus today, the only wild animal of any prominence in the country’s psyche is the agrino (Ovis orientalis ophion). This is a wild sheep, in the form of a subspecies found only on Cyprus – and evolved only on Cyprus. At face value, this form can be taken as properly indigenous to Cyprus, and an indication of the palaeofauna. However, this would be misguided, for everyone agrees that the agrino is descended from ancestors introduced by humans to Cyprus as recently as four thousand years ago (and certainly not longer than 12 thousand years ago).
 
So the big surprise of the western oceanic islands of the Mediterranean, viz the Gymnesics, is their caprine fauna. Quite unexpected, indeed unique, given that there’s not a trace of pigs or water buffaloes, and the status of geese and tortoises was nebulous and unremarkable (they occurred on these or nearby islands but there’s little evidence of their commonness or special adaptation on the Gymnesics).
 
On Cyprus, the real fauna consisted of hippo and elephant, both in diminutive form. There is enough fossil evidence to assume that geese were also common here, and perhaps the only important herbivores before the hippo and elephant arrived in the middle to late Pleistocene. As for tortoises, there is no sign that they have ever occurred naturally on Cyprus, although today the common tortoise (Testudo graeca iberia), a fairly large species of up to 35 cm and up to 2 kg, occurs there - certainly introduced by humans.
 
How can we interpret the surprising finding that it was hippo and elephant that occurred on Cyprus in the place of pig, water buffalo, and giant tortoise?
  
It’s true that both hippos and elephants swim excellently, and in that sense are preadapted to reach islands. But what this overlooks is that, once on the islands, they seem overspecialised relative to e.g. pigs, water buffaloes, geese and tortoises. Hippos and elephants are far from being ‘versatile and generalised herbivores’. So why have they succeeded on Cyprus where other candidates have failed, and why was the fauna of Cyprus so different from that of the Gymnesics?
  
(writing in progress)

Posted on 12 de julho de 2022, 08:48 AM by milewski