Diversity in Living Brains Cladistics

Having emphasized uniformities so forcefully, I must warn against underestimating the importance of diversity. All brains are different, and there are major differences both within and between species. Differences in brains within species are often difficult to measure with conventional anatomical and physiological methods, but since the brain is the control system for behavior, behavioral differences are evidence of differences among brains. Differences among species, of course, are much more dramatic.

The qualitative differences among the procyonids described by Welker could be presented as quantitative differences as well by measuring the amount of tissue in, for example, forepaw and rhinarial projections to the neocortex of procyonids. Differences among orders of mammals or classes of vertebrates are even more striking, but they too have not been quantified, perhaps because they are so obvious. And even when differences are great, they may be surprisingly difficult to describe quantitatively. One recognizes in an instant that the human brain is an unusual primate brain, for example, but the analysis of the relative size of its major parts usually shows it to be a perfectly normal primate (or mammalian) brain. In Fig. 2 there is nothing other than gross size to distinguish the human data, which fell on or near the regression lines determined for all the mammals in each sample. Also, as mentioned earlier, even the size of the prefrontal neocortex, often assumed to be uniquely important in human performance, is exactly as large as expected for a mammalian brain the size of a human brain. The uniqueness of human behavior is related to the size of the entire neural control system, and the correct conclusion about prefrontal executive function is that its size is appropriate to the very large brain systems that it controls.

A rigorous, though not really quantitative, analysis of the diversity of organization of the brain has been in the application of cladistic methodology. As indicated earlier, the results of this kind of analysis with brain features are essentially the same as those when other morphological features are used in the traits-by-species matrix that provides primary data for the analysis.

A cladistic methodology was applied by P.S. Ulins-ki, who took as his goal the reconstruction of probable features of the internal anatomy of brains at nodal evolutionary points in the evolution of reptiles, birds, and mammals. He first used the results of cladistic classification to determine nodes at which branching occurred when an ancestral species split into two daughter species. Second, he examined the brains of living representatives of the daughter species (or higher taxa). Finally, he constructed a hypothetical ancestral brain as a kind of lowest common denominator of the brains of the daughter species. The approach can be applied only to nodes in which surviving species from both branches exist. For example, taking living turtles and crocodiles to represent surviving species from the node of the early reptilian branching that led to these species, the ancestral brain can be constructed as having only those features that turtles and crocodiles share. With this procedure Ulinski could suggest various details about the ancestral brain of birds, crocodiles, lizards, and turtles. (From a cladistic perspective birds may be thought of as specialized reptiles derived from dinosaurs.) The nodal point in the history of the mammals, unfortunately (for this approach), is late in a "reptilian" synapsid lineage, represented today only by mammals. The mammal-reptile transition brain could be reconstructed only if synapsids at a reptilian grade of brain evolution had survived; but none have. The reconstruction of the brain at the reptile-mammal transition is, therefore, impossible using his procedure.

It would be possible to temper this conclusion, which is based on qualitative internal features of the brain, by analyzing the superficial anatomy using data on fossil endocasts, although such data are sparse for synapsids. Where they exist, in therapsids, they suggest a size pattern comparable to that in living reptiles. The transition from mammal-like reptiles to mammals is documented in the endocasts, and at present it seems to be reflected primarily as both enlargement of the brain (encephalization) and a major reduction in body size.

In his analysis of mammalian brain evolution, particularly the diversity of organization of somatosensory neocortex, J.I. Johnson presents an impressive catalog of detail on differences but concludes that "a great many features are constant across all mammals, from platypus to monkey, rat, cat, and sheep.'' The major variations "include [amount of] multiplication of representations of certain body parts" and details of the representations. He also notes few general trends of organization but comments that the appearance of "association cortex'' intercalated between somatosen-sory and visual neocortex is haphazard across species, and that its appearance seems "to have something to do with the use of limbs as information-gathering and manipulating organs.'' Johnson's conclusions are consistent with the notion that the pattern of organization of the brain in a species that differentiates it from other species follows no general principles in the mammals but is part of the specialization of each species. In cladistic analyses, Johnson and his associates found that the phylogenetic tree in mammals deduced from 15 brain traits was essentially the same as that deduced from other traits.

The quantification of diversity depends on the measurement of size. The evolution of brain size in mammals has led to the diversification that was already evident in the data of Fig. 2, with species having brains as small as 0.1 g (pygmy shrew) and as large as 8 kg (killer whale). These all evolved from a single species of mammal (according to the monophyly accepted today) that lived more than 200 Ma. I outline the history later, but we must first understand how the diversity of size is analyzed.

C. Allometry and Encephalization

Body size accounts for 80-90% of the variance in brain size between species, a relationship described by an allometric equation: the regression of the logarithms of brain size on body size. The distance of a species from the regression line is a measure of its encephalization. Because the scales are logarithmic, this distance, or residual, is an encephalization quotient—the ratio of actual brain size to expected brain size. Encephalization is a characteristic of a species; it is usually meaningless to discuss differences within a species in encephalization.

Allometry and encephalization do not have to be defined by regression equations and residuals, but most recent work on brain evolution involving brain body allometry uses this approach, which might be called ''parametric'' since it involves the estimation of the parameters of a normal probability distribution. Instead of the regression, the data can be described with minimum convex polygons enclosing the data points of the groups to be compared, but there are currently no quantitative methods to analyze the polygons.

Minimum convex polygons described the location of human and dolphin data in Fig. 2A, and the brain/ body data for the same insectivores, prosimians, and anthropoids as in Fig. 2B are graphed in Fig. 3, with polygons drawn around each group to compare them

101 102 103 104 105 Body Weight (grams)

Figure 3 Convex polygons to differentiate insectivore, prosimian, and anthropoid data on brain weight and body weight (from a chapter for Zaidel; see footnote on first page).

101 102 103 104 105 Body Weight (grams)

Figure 3 Convex polygons to differentiate insectivore, prosimian, and anthropoid data on brain weight and body weight (from a chapter for Zaidel; see footnote on first page).

with respect to encephalization. (Recall that Fig. 2B related the size of the cerebellum to the size of the whole brain; in Fig. 3 the relationship is of the whole brain to the size of the body.) It is not difficult to distinguish relative brain size among the groups since there is little overlap. All the polygons are oriented upward. There is slight overlap between the insectivores and prosimians and a bit more overlap between prosimians and anthropoids. From Fig. 3, one would describe the order of encephalization of these groups as follows: insectivores are least encephalized; prosi-mians are intermediate, and anthropoids are most encephalized. These data are also described by regression equations in Fig. 4.

The work by Stephan's group is especially relevant for evolutionary analysis because of the species they used. They worked with insectivores to represent a primitive grade of brain evolution and to provide an evolutionary perspective on the human brain. The issues are more complex, of course, but insectivores are reasonable models for the base group from which most placental species evolved. They resemble the earliest mammals both skeletally and in their endocasts. Although primates are currently a highly encephalized order of mammals, they are also a very ancient order, probably derived during the Late Cretaceous period from a species comparable to living insectivores or tree shrews. Comparisons between insectivores and primates are thus very appropriate for our topic.

Figure 4 Regression analysis of the data shown in Fig. 3. Some species are named to indicate diversity of sample. (A) Separate regressions and correlation coefficients for the three groups: insectivores: Y = 0.05 X0 67, r = 0.946; prosimians: Y = 0.14 X0 66, r = 0.960; anthropoids: Y = 0.13 X0 75, r = 0.972. (B) Lumping the data for an overall regression for all 76 species: Y = 0.05 X0 91, r = 0.966. (redrawn with permission from Jerison, 1991).

Figure 4 Regression analysis of the data shown in Fig. 3. Some species are named to indicate diversity of sample. (A) Separate regressions and correlation coefficients for the three groups: insectivores: Y = 0.05 X0 67, r = 0.946; prosimians: Y = 0.14 X0 66, r = 0.960; anthropoids: Y = 0.13 X0 75, r = 0.972. (B) Lumping the data for an overall regression for all 76 species: Y = 0.05 X0 91, r = 0.966. (redrawn with permission from Jerison, 1991).

D. A Bit of Theory

Issues in parametric quantification of encephalization as they apply to insectivores and primates are suggested in Fig. 4. The two graphs present the same data, fitted by straight lines in different ways. Fig. 4A shows the regression of log brain size on log body size computed separately for the three groups; Fig. 4B is a single regression for all 76 species. The three regression lines in Fig. 4A provide the same information as the polygons in Fig. 3. But if one is interested in curve fitting all the regression lines fit remarkably well (r>0.94) despite their different slopes. These slopes on log-log axes are the exponents of the equations written as power functions, and the value of a ''true exponent'' has been the subject of considerable debate during the past decade. This is where a little theory may help.

The emerging consensus is that an exponent of 3/4 is the correct value. I have quarreled with this view, arguing in favor of a 2/3 exponent, which has theoretical significance for dimensional analysis of the brain's work in mapping information from the external environment. It is true that empirical analyses of large enough samples of species, or of properly sampled groups of species, lead to the 3/4 exponent when the fit is statistical, but I believe that the theoretical value of 2/3 is nevertheless correct. The point is that the 2/3 value is required by the dimensional problem in order to convert data about a surface into data about a volume (a ''mapping''). It reflects the fact that our information about the external world is spread across a surface consisting of sensory cells distributed throughout the body (skin, retina, organ of corti, olfactory epithelium, etc.) and that information is pumped up to neurons distributed through a kind of conceptual surface in a brain. I have assumed a fixed cortical thickness as representing that brain surface, and that the measure of brain volume in brainbody allometry is converted into a measure of that surface area. However, since the conversion is by a physical system that takes up space, one has to take into account the thickness of the map formed by the cortical ''surface.'' To explain the difference between a 2/3 exponent required for the mapping and the 3/4 exponent found empirically, I have argued that this thickness as estimated by the thickness of neocortex is known to be greater in larger brains, varying approximately with the 1/9 power of body size. The value 3/4 is approximately the sum of 2/3 + 1/9. The theoretical value of 2/3, which is meaningful for the brain's mapping function, thus leads to an expected empirical value of 3/4.

E. Encephalization

The fact of encephalization is evident in the vertical displacement of the lines that are fitted to the three groups (Fig. 4A) or of the minimum convex polygons

(Fig. 3). Since the polygons do not require the dubious assumptions of statistical curve fitting, they may be preferred for describing encephalization. They are certainly preferred if they are adequate for answering questions about whether groups are equal or differ in encephalization.

The degree of encephalization in living vertebrates is summarized in Fig. 5. The polygons enclose all the available data on the indicated classes. The data were assembled from a variety of sources. The main inference from Fig. 5 is that one can characterize birds and mammals jointly as "higher" vertebrates, and reptiles, bony fish, and amphibians can be characterized as "lower" vertebrates. The polygons do an adequate job, although the addition of data on cartilaginous fish (sharks, rays, and skates), jawless fish (agnathans), and electric fish (bony fish: Mormy-riformes) makes it difficult to distinguish the groups by inspection. The additions are of relatively few species. As mentioned previously, the present consensus recognizes about 50,000 vertebrate species. Of these, to the nearest thousand, about 25,000 are bony fish, 6000 are reptiles, 4000 are amphibians, 10,000 are birds, and 5000 are mammals. There are about 800 cartilaginous fish species and 70 agnathan species.

From the evidence presented earlier in Fig. 2A, it is appropriate to assume that the amount of encephali-zation measures the information processing capacity of a brain, adjusted for body size. It is therefore rrm]—I I Uini|—I 11 IIIII|

uni—I I I mill— I—I I I mill—I I I mill—i i mini_i i mini_i I_■ I

<10-1 10° 101 102 103 104 105 106 107 108 Body Size (Grams)

Figure 5 Brainbody relations in 2019 living vertebrate species enclosed in minimum convex polygons. The samples are 647 mammals, 180 birds, 1027 bony fish, 41 amphibians, 59 reptiles, 59 chondrichthyans (sharks, rays, and skates), and 6 agnathan fish (redrawn with permission from Jerison, Roth and Wulliman, 2001).

uni—I I I mill— I—I I I mill—I I I mill—i i mini_i i mini_i I_■ I

<10-1 10° 101 102 103 104 105 106 107 108 Body Size (Grams)

Figure 5 Brainbody relations in 2019 living vertebrate species enclosed in minimum convex polygons. The samples are 647 mammals, 180 birds, 1027 bony fish, 41 amphibians, 59 reptiles, 59 chondrichthyans (sharks, rays, and skates), and 6 agnathan fish (redrawn with permission from Jerison, Roth and Wulliman, 2001).

appropriate to consider the ecological requirements met by increments in processing capacity in different groups of species. Not much debate is required to see mammals and birds as higher vertebrates in this regard, given the normal complexity and plasticity of behavior observed in these groups. No reasonable speculations have been offered for the position of the cartilaginous fish as overlapping the higher and lower groups, but the place of electric fish reflects an unusually enlarged cerebellum in these species, related to processing information from their electric organs. It is unclear why that processing should require as great an investment in neural machinery. The position of jawless fish has been placed below the bony fish polygon, leading to speculations that there may have been a reduction in brain size related to the parasitic habits common in this group, particularly among lampreys. However, as evident from Fig. 5, agnathans, though relatively small-brained, fall more or less within the fish polygon, making such speculations unnecessary.

The approach signaled by Fig. 5 enables us to evaluate fossil endocasts with respect to encephaliza-tion, providing a direct evolutionary window to the patterns of change that led to the current diversity in brain size. I present such a nonparametric analysis as well as a parametric (regression) analysis of neocorti-calization in the next section. In the analysis by convex polygons, I will be concerned with the evolution of birds and mammals from the reptiles and the utility of the method for some conclusions about dinosaur brains.

V. QUANTITATIVE ANALYSIS: FOSSILS A. Vertebrate History

Vertebrates first appeared during the past 500 million years of the earth's 4.5 billion (4.5 x 109) year existence, and Table II provides a synopsis of their history. Here are some points to remember.

First, the world was very different in the distant past compared to the present. During the Paleozoic Era, there were times when there was only a single global continent (Pangea), but landmasses joined and separated with the passage of time. The global map was significantly different during the Mesozoic, with major masses (Gondwanaland and Laurasia) during the Paleozoic and Mesozoic Eras. There were warmer and more stable climates during the Mesozoic, and the continents were drifting toward their present loca

Table II

Synopsis of Vertebrate Evolution

Table II

Synopsis of Vertebrate Evolution

Era

Period and epoch

Age (years x 106)

Fauna (first appearance)

Cenozoic

Quaternary

Conquering Fear In The 21th Century

Conquering Fear In The 21th Century

The Ultimate Guide To Overcoming Fear And Getting Breakthroughs. Fear is without doubt among the strongest and most influential emotional responses we have, and it may act as both a protective and destructive force depending upon the situation.

Get My Free Ebook


Post a comment