aat

PARSIMONY OF AQUATIC AND TERRESTRIALHYPOTHESES:

HOW MANY HYPOTHESES DO WE NEED?

John H. Langdon

University of Indianapolis

langdon@uindy.edu

Prepared for

Water and Human Evolution:

Savanna, forest or aquatic origins of our hominid ancestors?

Friday, April 30th, 1999

Ghent, Belguim

The aquatic ape hypothesis (AAH) has been supported primarily fromtwo lines of argument. First, there is a long list of unusual humananatomical features that appear to be consistent with aquaticadaptation. Secondly, it is argued that the hypothesis is moreparsimonious than competing terrestrial models for explaining theevolved human condition.

Parsimony does not necessarily prove one hypothesis over another;nonetheless, it is an important quality of an argument in science andleads us to favor one hypothesis over one less parsimonious. I haveargued elsewhere that the structure of adaptationist arguments inhuman evolution require separate hypotheses to explain individualtraits and that an umbrella hypothesis that seems to encompass alarge number of traits within a single explanatory framework does notnecessarily increase parsimony (Langdon 1997). The AAH, inparticular, is no more parsimonious in its treatment of individualanatomical features than available terrestrial models. Furthermorethe underlying premise of major shifts in habitat is inherentlyunparsimonious.

This paper continues that argument in comparing the structure ofaquatic and terrestrial hypotheses. I argue that the large number ofanatomical characters in question can be reduced to a smaller numberof important adaptive problems. Most of the important questions aboutthe evolution of the modern human body may be summed up in five suchadaptive complexes. In these issues, the aquatic hypothesis is notmore parsimonious. In fact, for specific explanations, the competingaquatic and terrestrial models face the same problems.

FAT, SWEAT, AND NAKED SKIN

Why are whales hairless?

In order to build the argument that hairless humans were aquatic,it is not enough to point out that many aquatic mammals are hairless.One must ask whether they are hairless for the same reasons. Inaddition to aquatic or semiaquatic mammals such as whales, porpoises,and hippos, terrestrial mammals including rhinoceros, elephants, andnaked mole rats are also hairless. Making sense out of hair lossmeans understanding the functional and selection factorsinvolved.

Are aquatic mammals functionally hairless? Fully aquatic mammalsare hairless, but "semiaquatic" mammals generally are not (Table1).

• All cetaceans are hairless. Presumably the state ofconstant immersion in water negates the insulating properties ofhair.

• Among the pinnipeds (seals and their relatives), who aresufficiently committed to water that they have become inefficient onland, only the three largest genera are mostly hairless &emdash;Odobenus, walrus; Eumetopias, Steller's sea lion; and Mirounga,elephant seal.

• Among 18 genera of quadrupeds that spend significantamounts of time in the water, including wading while foraging, onlytwo genera of hippopotamus are hairless. The water buffalo and tapirhave sparse hair. The others have well developed coats of fur.

If human hairlessness is evidence of aquatic adaptation, then thisdata tells us our ancestors spent a period in which they never cameto land. They were expressly not semiaquatic. Hair appears to be animportant adaptation for animals who move between dry and wetconditions.

On the other hand, there are 17 genera of terrestrial mammals withloss or reduction of hair (Table 2). Six ofthese are elephants and rhinoceros &emdash; representing six of theeight largest land mammals. Four are related genera of pigs; four arethe clade of great apes and humans. The pangolin and giant armadilloare armored, presenting a unique situation. The naked mole rat is asmall mammal that spends its life in underground burrows.

Body hair functions in most mammals as insulation to retain bodyheat. It also protects the skin from parasites and is part of asensory organ. When it has been lost, it is because one or more ofthose functions are no longer effective or appropriate.

The most common circumstance in which body hair is lost is with alarge body size (greater than 1000 kg). Increased body mass, with adecreasing relative surface area, makes it difficult to shed the heatgenerated by normal metabolism, including gut fermentation. The largemammals that are exceptions to this pattern include the giraffe, withan anomalously large surface area, and the American buffalo, eland,and yak &emdash; all at the lower limit of this range (Table3). In large mammals, loss of body hair is probably an adaptationto reduce the insulation retaining body heat. Many of these on theland will further cool themselves off by wallowing in water, mud, ordust.

A second circumstance that favors the loss of body hair isconstant immersion in water. Water is a natural enemy of endothermicanimals, since large bodies of water are nearly always colder thanthe body temperature. Many mammals and birds can maintain an oiledcoat of hair or feathers that are waterproof up to a point. As longas they can trap air pockets against the skin, such coats continue tofunction as insulation. They are especially important when the animalcomes to land and dries in the air. If the animals stays in thewater, however, eventually the hair becomes saturated and looses itsability to insulate. Such animals have all lost body hair and relyprimarily on fat for insulation.

A few hairless mammals are fossorial (burrowing). Two of these&emdash; the giant armadillo and the pangolin &emdash; are alsocovered with armored scales that partially preclude hair. A third isthe naked mole rat.

We thus can recognize at least three circumstances that favor hairreduction: constant immersion in water, gigantism, and possiblyburrowing. Hominoids and pigs do not match any of these three and mayrepresent a fourth habitat that favors hair reduction &emdash; ashaded tropical niche, such as a forest.

Have humans lost hair for the same reason as whales? If thatreason is stated in terms of aquatic lifestyle, the answer is no. Ifthat reason is stated as the practical loss of insulative properties,the answer is probably yes. In order to investigate this further, weneed to understand better the patterns of hair reduction. Forexample, many mammals have two different layers of hair, under furand guard hairs, that serve different functions. When other mammalsare dichotomized as either naked or hairy, the distinctions betweenthese layers is lost, though it is functionally significant.

Why are we fat?

Humans are claimed to have more fat cells and a greater percentageof body fat. To the extent that this is true, our subcutaneous fatplays a role in an unusual metabolic strategy that maintains a highenergy flow to develop and feed the brain and supports prolongedbouts of exercise. A key element of this is a thermoregulatorystrategy that uses a combination of hairlessness, sweating, andcutaneous vasodilation to rid the body of excess heat. This strategyrelates a number of unusual human traits that cannot be understood inisolation, but do make sense together:

wide distribution and proliferation of eccrine glands
both systemic and regional thermally sensitive controls of eccrine glands
profligate expenditure of water during sweating with possibility of dehydration
increased subcutaneous fat
reduction of body hair
proliferation of cutaneous vasculature
neural control of cutaneous vasculature for thermoregulation
expenditure of salt during sweating

However, fat is a problematic characteristic. The amount of adultbody fat varies greatly in different populations due to a combinationof genetic, behavioral, and economic factors. Which population, ifany, can be considered characteristic of our species? Certainly notmiddle-class North Americans. Is subcutaneous fat sufficientlywidespread across the body to be an effective insulation? Pond (1991)indicates that it exists in the same regions of the body as in othermammals, though somewhat expanded in area. Many areas of skin aretherefore without such protection.

Additional functions of fat include storage of energy and oflipid-soluble molecules such as estrogens. Fat therefore represents anutritional buffer in a variable environment. This may be especiallyimportant in infants, where high levels of nutrients are needed bythe growing brain. Indeed, humans infants to possess a strikingpercentage of body fat.

The AAH explains the unusual combination of hairless and fat as aparallel with many aquatic mammals, thus indicating a similar aquaticadaptation for insulation and/or buoyancy. Alternatively,understanding these are two of many traits functionally integratedinto thermoregulation brings us to a single adaptive problem: Why dowe have this unique thermoregulatory strategy? Hypotheses areavailable (e.g., Langdon and Nawrocki 1997).

It is not difficult to recreate a gradual evolution of thiscomplex. None of these traits appear de novo. Proliferation ofeccrine glands and relative hair reduction are shared with livinghominoids, though humans carry these traits further (Montagna 1972).The fat reserves are mere expansions of deposits found in otherspecies. The neural control makes sense only as a thermoregulatoryadaptation and not for salt balance. There is no great evolutionaryleap that requires an extraordinary explanation.

BIPEDALISM

Why are we bipedal?

Bipedalism has arisen several times &emdash; e.g., in dinosaurs,kangaroos, kangaroo rats, and possibly birds. Again there aredifferent stories for different taxa, but none of them involve water.Bipedalism in hominids has driven a large number of anatomical andbehavioral changes. Some of these are:

forward positioning of foramen magnum
barrel-shaped ribcage
shortened lumbar spine
inflation of vertebral bodies and lower limb bones for shock absorption
shorter, broader pelvis
alignment of sacroiliac and acetabular joints
broader, wedge-shaped sacrum
full hip and knee extension
locking mechanisms of hip and knee
enlargement of gluteus maximus and assignment as extensor
lower limb elongation
carrying angle of femur
elongated foot
short toes
adducted and elongated hallux
breathing independent of respiration
enlargement of semicircular canals
frontal sex
vaginal depth
hands freer for carrying food, babies, and tools

The AAH proposes that bipedalism is an adaptation for wading thatsecondarily adapted us for swimming. Independent hypotheses areproposed to account for increased vaginal depth, frontal sex, andindependent respiratory control (Morgan 1990). If we recognize thatmost, if not all, of these traits relate to the single adaptation ofbipedalism, we can reduce them to a single evolutionary problem: Whydid we become bipedal?

Obviously at some point our ancestors shifted from an arborealclimbing niche to a non arboreal bipedal one. The AAH explains thisas a shift from arboreal resources to aquatic resources. Terrestrialmodels suggest a shift from arboreal resources to ground resources.There is little difference between them at this point.

Again, the evolution of bipedalism required only a small amount oftinkering at any one time. Primates are all facultatively bipedal.Hominoids in particular commonly position the trunk in uprightpositions and support body weight on the lower limbs when climbing.Contrary to the arguments of the AAT (e.g., Morgan 1990), obligatebipedalism is a small step away.

BREATHING AND SPEAKING

Why can people talk?

Speech is a unique human attribute, although a great facility forcomplex vocalizations can be observed in some marine mammals and someterrestrial birds. There is no obvious tie between speech and anaquatic habitat.

Speech has several anatomical requirements:

fine motor control of voluntary breathing
descended pharyx
fine motor control of the tongue and larynx
appropriate language processing centers in the brain

It is not at all clear when these developed. Chimpanzees have ahomologue of Wernicke's area (Gannon, et al. 1998), but are presumedto lack the specialized neural development for language acquisition.Control over breathing presupposes only bipedalism. The narrow spinalcanal in H. ergaster has been interpreted as evidence againstthe neural control necessary for speech (MacLarnon 1993).Paleontologists disagree over interpretations of the pharynx inarchaic Homo. The full set of anatomical correlates of speechprobably evolved only in the past few hundred thousand years,although it is difficult to infer from that exactly when languagecommunication began.

How does the AAH explain these? It proposes four independenthypotheses.

H1: Breath control is an adaptation for diving.
H2: The descended pharynx is an adaptation for rapid intake of oxygen following a dive.
H3: Fine motor control of the tongue and larynx is something aquatic animals, such as dolphins, evolved for reasons not specified.
H4: Human language evolved on the back of this anatomical substrate (for social reasons?).

Alternatively, the anatomical design may have been driven by theadvantages of language. This demands an answer only to one question:Why did speech evolve?

HANDS AND TOOLS

Why do humans use tools?

Since the 1960's, we have been fascinated and amused by thediscovery of a variety of animals that use or even make tools in thewild. The list includes chimpanzees, orangutans, capuchins, otters,elephants, Egyptian vultures, Galapagos finches, New Caledoniancrows, and insects.

Human dexterity with tools also depend on a number of anatomicalcorrelates:

opposible thumb
fine, independent finger movements
nails rather than claws
motor control of hands, in particular

We are indebted to our primate heritage for most of these.Fine-tuning motor control was a late addition. Given the facility ofour closest hominoid relatives for manipulation and the example ofchimpanzee culture, our ancestors needed only a reason to usetools.

The AAH proposed that tools were used to process shellfish, byanalogy with otters. Terrestrial models can point chimpanzees toargue that tools were equally useful for opening nuts. Our question:What use of tools was of sufficient importance to drive selection forcultural abilities?

BRAINS

Why don't all species have large brains?

This question is central to human evolution. The problem is notthat there are no good answers, but that there are many and none haveoverwhelming evidence. Correlates of brain expansion include:

large brains
increased intellectual capacity
language
complex social behavior
slow postnatal growth
long period of lactation
long pregnancy
single births
high birth weight
late maturation
great longevity
large body size
vascular arrangements to cool the brain

The AAH has offered the challenges of shifting between terrestrialand aquatic habitats not once, but twice, as the challenge thatspurred brain growth. Recent terrestrial models have pointed to thefluctations in environmental conditions over the past five millionyears as challenging cultural development (e.g., Potts 1996). Thesetwo models do not differ substantively on this problem.

HOW MANY HYPOTHESES DO WE NEED?

I have attempted to consolidate the list of anatomical traits intoa few important adaptive complexes that distinguish humans. It is theadaptive complexes, not a numerical list of traits, that shouldcommand our attention. Some odd human characters remain unexplained,but do not obviously relate to aquatic adaptations (e.g., "psychictears," milk composition, and breasts). We should be suspicious,however, of any model that readily explains everything. Probably someof the different sets of adaptations can are the result of unrelatedselective pressures.

There are five critical questions that, if answered, go far toexplaining human evolution.

1. Why (as a result of what selective pressures) did we shift fromarboreal climbing to terrestrial bipedalism?

2. Why did we assume our unique thermoregulatory physiology?

3. Why did speech evolve?

4. Why do we make better tools than do other animals?

5. Why do we have a large brain?

The answers to these questions are the same, whether they comefrom terrestrial or aquatic models. The best current answer to eachquestion is not "I don't know." Rather it is "I don't know, but Ihave an idea." The AAH does not produce answers that are moreparsimonious or better supported than those of terrestrial models.There are no extraordinary evolutionary leaps involved that requireextraordinary explanations.

TIMING

Could all of these selective complexes have been the product of adiscreet aquatic phase in our evolutionary past? To answer that, wemust ask when did this period of selection did occur? During whichphase of human evolution were hominids committed to an aquaticniche?

The model makes two predictions. Relevant anatomical traits (thosewhich evolved because they were adaptive in an aquatic niche)appeared during this period of aquatic selection. Also, hominidsduring this period were not less adapted to water than are modernhumans. With these in mind, we can turn to the fossil and comparativerecord. Not all of the anatomical evidence cited to support the AAHcan be expected to show up in the fossil record, but a number can beplaced in an appropriate chronology (Table4).

Of all the anatomical traits identifiable in fossil hominids thatare argued to represent aquatic adaptation, only one is not expressedas well in modern humans as in extinct populations. Verhaegen (1993)has suggested that the great mass and density of bone in theposterior cranium and the cortex of limb bones of Homo erectuswere adaptations to make it easier to dive. I note this is oddlycontrary to the notion of fat as needed to increase bouyancy. Othertraits claimed to be associated with the aquatic niche appearindependently at different times throughout hominid history.

Some traits show evidence of selection in more than one timeperiod. The brain has expanded continually thoughout the past 3.0million years (Figure 1). Brain evolution has thus been a gradualprocess, representing continuous or repeated selective forces at workthroughout the hominid lineage. Bipedalism evolved in a least twospurts. The earliest australopithecines are bipedal by 4.0 Mya, whileHomo ergaster shows a modernization of the postcranium formore effective bipedalism after 2.0 Myr. Elements relating to breathcontrol and speech appear at different times, as well.

FIGURE 1. FOSSIL HOMINID BRAIN SIZE THROUGH TIME.

Each symbol represents the endocranial volume of a fossil hominidspecimen.

black squares: Australopithecus

red diamonds: Homo

Vertical scale = cranial capacity (c.c.)

Horizontal scale = millions of years

Therefore, if all of these elements of our anatomy really doreflect aquatic adaptation, such selection must have been acting onhominids continually from at least the Late Miocene to the LatePleistocene, if not the present. This is a more extreme statement ofthe AAH than Morgan has discussed, but corresponds to ideas thatVerhaegen has promoted. It suggests that humans today are as aquaticas hominids have ever been. Homo sapiens is the aquaticape.

How complete is our committment to the water? The answers areinferred from published argument.

• Our committment is complete enough that it has lead tosubstantial reshaping of the body for wading, swimming, anddiving.

• Our committment is complete enough that it has lead to aloss of traits useful on land but without function in the water. Forexample, our sense of smell is much diminished presumably because weno longer spend enough time out of the water to use it.

• Our committment is complete enough that it has lead toanatomical changes that were maladaptive on land. For example, duringthe evolution of bipedal posture and locomotion our body mass wasconsistently supported by water so that mechanical forces that wouldhave discouraged the assumption of bipedalism on land werenegated.

• Our committment is complete enough that reproduction andchild care, if not childbirth itself, occurred in the water. Thusestrus was lost because olfactory signals were irrelevant; childbirthoccurs easily in the water; breasts evolved to suckle children in thewater; and women have long hair to give young children something tohold onto.

Unfortunately, these statements do not describe the behavior ofany modern humans, many of whom live in desert conditions or whomnever interact with significant bodies of water. Certainly humansexploit aquatic resources for food, drink, and recreation. However,there is not sufficient evidence to suggest that our relationshipwith an aquatic niche has influenced the course of humanevolution.

TABLE 1. HAIRLESSNESSAND AQUATIC HABITAT IN MAMMALS

Fully aquatic mammals

Cetacea &emdash; all	whales, dolphins and relatives	hairless
Sirenia &emdash; all	manatees and relatives	hairless

Mostly aquatic mammals &emdash; pinnipeds (in order of size)

Mirounga	elephant seal	variable hair	3500 kg
Odobenus	walrus	sparse hair	1700
Eumetopias	Steller's sea lion	sparse	1100
Hydrurga	leopard seal	furred	500
Leptonychotes	Weddell seal	furred	450
Cystophora	hooded seal	furred	400
Zalophus	California sea lion	furred	400
Arctocephalus	southern fur seal	furred	363
Erignathus	bearded seal	furred	360
Otaria	South American sea lion	furred	350
Halichoerus	gray seal	furred	310
Neophoca	Australian sea lion	furred	300
Monachus	monk seal	furred	300
Lobodon	crab-eater seal	furred	300
Ommatophoca	ross seal	furred	216
Phoca	ringed seal	furred	148
Callorhinus	northern fur seal	furred	50

Highly aquatic Quadrupeds

Hippopotamus	hippopotamus	hairless	4500 kf.
Thalarctos	polar bear	furred	800
Choeropsis	pygmy hippopotamus	hairless	270
Enhydra	sea otter	furred	45
Pteronura	giant otter	furred	34
Aonyx	clawless otter	furred	34
Castor	beaver	furred	32
Lutra	river otter	furred	14
Mustela	American mink	furred	2

Wading and aquatic browsing mammals

Bubalus	water buffalo	sparse	1200 kg.
Alces	moose	furred	825
Kobus	waterbuck	furred	500
Tapirus	tapir	sparse	320
Blastocerus	swamp deer	furred	150
Hydrochoerus	capybara	furred	50
Hydropotes	water deer	furred	30
Nasalis	proboscis monkey	furred	22
Cynogale	otter civet	furred	5

Body mass represents top of normal range. Data from Nowak 1991 andParker 1990.

TABLE 2. HAIRLESSNESS INTERRESTRIAL MAMMALS.

Mammals lacking normal body covering (in order of size):

Loxodonta	African elephant	hairless	6300 kg
Elephas	Asian elephant	hairless	5400
Rhinoceros	Indian rhino	hairless	2200
Ceratotherium	white rhino	hairless	3600
Diceros	African rhino	hairless	1800
Dicerorhinus	Asiatic rhino	hairless	1000
Sus	wild boar	sparse	350
Hylochoerus	forest hog	sparse	275
Gorilla	gorilla	"often sparse"	275
Phacochoerus	wart hog	sparse	150
Babyrousa	babirusa	hairless	100
Pongo	orangutan	"never thick"	90
Homo	human	hairless	75
Pan	chimpanzee	sparse	70
Priodontes	giant armadillo	hairless	60
Manis	pangolin	hairless	33
Heterocephalus	naked mole rat	hairless	.08

Body mass represents top of normal range. Data from Nowak 1991 andParker 1990.

TABLE 3. LARGETERRESTRIAL MAMMALS (at least 900 kg) WITH FULL HAIR

Giraffa	giraffe	1930 kg.
Bison	American buffalo	1000
Bos	yak	1000
Taurotragus	eland	1000
Synceros	African buffalo	900

Body mass represents top of normal range. Data from Nowak 1991 andParker 1990.

TABLE 4.CHRONOLOGICAL APPEARANCE OF ANATOMICAL TRAITS ASSOCIATED WITH ANAQUATIC NICHE.

Time period

Traits

Evidence

Aquatic Function

pre-Miocene (primate ancestry)

dextrous hands

shared character

aquatic feeding

by Late Miocene (last common ancestor)

rudimentary tool use

encephalization to ape grade

shared character

shellfish feeding

habitat shift

Pliocene (Australopithecines)

bipedalism

plantigrady

toes shortened

toes subequal

adducted hallux

elongated foot

hip extension

?decreased olfaction

paranasal sinuses

fossils, footprints

some fossils

facial shortening

fossils

wading or swimming

swimming

wading

swimming

(non-adaptive)

buoyancy

Late Pliocene (early Homo)

stone tools

encephalization beyond 600 cc

archaeological record

fossil Homo

shellfish feeding

habitat shift

Pleistocene (Homo erectus grade)

refinement of bipedalism

limb elongation

toe shortening

barrel thorax

descended larynx?

external nose

dense occipital, limb bones

encephalization

fossils

ribcage structure

basicranial anatomy

facial anatomy

fossils

wading

swimming

diving

swimming/diving

swimming, buoyancy control

habitat shift

Late Pleistocene (archaic and modern Homo)

motor control of diaphragm?

encephalization

vertebral canal

fossils

diving

habitat shift

REFERENCES

Gannon, Patrick J., Ralph L. Holloway, Douglas C. Broadfield, andAllen R. Braun 1998. Asymmetry of chimpanzee planum temporale:humanlike pattern of Wernicke's brain language area homolog.Science 279:220-222.

Langdon, John H. 1997. Monolithic hypotheses and parsimony inhuman evolution: a critique of the aquatic ape hypothesis. Journalof Human Evolution 33:479-494.

Langdon, John H., and Stephen P Nawrocki 1997. The evolution ofendurance &emdash; toward a synthesis of skeletal and soft tissueevolution. Ms. presented at the Paleoanthropology Society AnnualMeeting, April 1997.

MacLarnon, Ann 1993. The vertebral canal. In Alan Walker andRichard Leakey, eds., The Nariokotome Homo erectusSkeleton. Cambridge, MA: Harvard University Press. Pp.359-390.

Montagna, William 1972. The skin of nonhuman primates. AmericanZoologist 12:109-124.

Morgan, Elaine 1990. The Scars of Evolution. New York:Oxford University Press.

Nowak, Ronald M. 1991. Walker's Mammals of the World, 5thed. Baltimore: Johns Hopkins University Press.

Parker, Sybil P., ed. 1990. Grizmek's Encyclopedia of Mammals. NY:McGraw Hill.

Pond, Caroline M. 1991. Adipose tissue in human evolution. InMachteld Roede, Jan Wind, John M. Patrick, and Vernon Reynolds, eds.,The Aquatic Ape: Fact or Fiction? London: Souvenir Press. Pp.193-220.

Potts, Richard 1996. Humanity's Descent. NY: AvonBooks.

Verhaegen, Marc 1993. Aquatic versus savanna: comparative andpaleo-environmental evidence. Nutrition and Health9:165-191.

Home