Research Article
Constructional morphology of sand dollars
- Adolf Seilacher
-
- Published online by Cambridge University Press:
- 08 February 2016, pp. 191-221
-
- Article
- Export citation
-
This paper analyzes an aberrant group of echinoids in terms of constructional morphology, i.e., as modification of an established “Bauplan” by a set of new functional and morphogenetic constraints and possibilities. The characteristics of sand dollars (flat test, spine differentiation, branched food grooves, lunules) are related to a particular combination of burrowing and sieve feeding in sandy sediments. It has independently evolved from less specialized Clypeasteroids in at least three lineages (Scutellina, Rotulidae, Arachnoididae), which have solved inherent problems differently (sutural interlocking; growth patterning of food grooves and canal systems; lunule formation; weight belts). These three groups have radiated in different degrees due to their different palegeographic histories.
Role of ideas in advancing paleontology
- Ingrid M. Krohn
-
- Published online by Cambridge University Press:
- 08 February 2016, pp. 67-76
-
- Article
- Export citation
-
Modern biological principles and mathematical techniques applied to fossil data may contribute to a rigorous theoretical framework for the analysis of macroevolutionary trends. Such cross-fertilization may also serve to correct interpretative biases common in paleontological literature.
Community paleoecology as an epiphenomenal science
- Antoni Hoffman
-
- Published online by Cambridge University Press:
- 08 February 2016, pp. 357-379
-
- Article
- Export citation
-
The most important topics discussed thus far by community paleoecologists are: community approach to paleoenvironmental reconstruction, community development in evolutionary time, and community constraints upon species evolution. The first step in community-paleoecological analysis is to reconstruct the paleocommunity or paleoecosystem. However there are severe methodological limitations to any inference drawn from the composition and structure of a fossil assemblage. These result from various taphonomic biases. These constraints upon reliability of a community-paleoecological analysis are the least severe in the case of Cenozoic (and possibly Cretaceous) tropical to subtropical, molluskdominated, subtidal, benthic communities.
The basic assumption of community paleoecology is that communities or biocoenoses represent a distinct, real level of biotic organization achieved through ecological integration of and coevolution among the species. This assumption seems to be invalid for two reasons. (1) The actual degree of community integration is in general insufficient to induce any driving forces for a structural development as predicted by the system theory. (2) The concept of biological reality and distinctness of the community level of biotic organization implies assignment of a significant role to group selection. The assumption that ecological communities achieve with time an equilibrium state representing an optimum habitat partitioning among the component species is invalid, at least as a generalization. This idea is largely falsified and refuted by a variety of ecological studies. Ecological communities are then merely an epiphenomenon of the overlap in distributional patterns of various organisms. There is no intrinsic, biotic mechanism inducing community dynamics in either ecological, or evolutionary time.
In spite of this conclusion, the so-called “community approach” under favorable taphonomic conditions is among the most reliable methods of paleoenvironmental reconstruction, and the data on so-called “community evolution” are relevant to the problem of a relationship between actually realized niche patterns and their environmental framework. The latter problem can also be approached through analysis of longevity-frequency-distributions of chronospecies found living together in various habitats. A preliminary investigation may indicate that the precondition to optimization of niche dimensions through coevolution among ecologically related species was met in subtidal benthic habitats but not in pelagic ones. To account for a niche subdivision among planktonic organisms, one has therefore to invoke a stochastic pattern of speciation rather than coevolutionary mechanisms.
Dinosaur eggs: gas conductance through the shell, water loss during incubation and clutch size
- Roger S. Seymour
-
- Published online by Cambridge University Press:
- 08 February 2016, pp. 1-11
-
- Article
- Export citation
-
The conductance of water vapor and respiratory gases by diffusion through the eggshells of Upper Cretaceous dinosaurs has been estimated from measurements of shell and pore geometry in fossil specimens. When compared to recent reptile and bird eggs for which nest environments are known, the highly porous eggshells of three dinosaur species indicate that the dinosaur nests were high in humidity and probably low in oxygen and high in carbon dioxide. Such conditions most likely occurred underground or within an incubation mound.
By isolating the eggs from the atmosphere, however, some large sauropods may have been forced to limit their clutch size to numbers small enough to prevent depletion of oxygen and elevation of carbon dioxide to intolerable levels in the nest. Fossil evidence supports this and suggests that one sauropod actually divided her large eggs into several clutches. Each small clutch probably had a metabolic rate similar to those of clutches produced by recent reptiles and mound nesting birds.
A kinetic model of Phanerozoic taxonomic diversity II. Early Phanerozoic families and multiple equilibria
- J. John Sepkoski, Jr.
-
- Published online by Cambridge University Press:
- 08 February 2016, pp. 222-251
-
- Article
- Export citation
-
The kinetic model of taxonomic diversity predicts that the long-term diversification of taxa within any large and essentially closed ecological system should approximate a logistic process controlled by changes in origination and extinction rates with changing numbers of taxa. This model is tested with a new compilation of numbers of metazoan families known from Paleozoic stages (including stage-level subdivisions of the Cambrian). These data indicate the occurrence of two intervals of logistic diversification within the Paleozoic. The first interval, spanning the Vendian and Cambrian, includes an approximately exponential increase in families across the Precambrian-Cambrian Boundary and a “pseudo-equilibrium” through the Middle and Late Cambrian, caused by diversity-dependent decrease in origination rate and increase in extinction rate. The second interval begins with a rapid re-diversification in the Ordovician, which leads to a tripling of familial diversity during a span of 50 Myr; by the end of the Ordovician diversity attains a new dynamic equilibrium that is maintained, except for several extinction events, for nearly 200 Myr until near the end of the Paleozoic. A “two-phase” kinetic model is constructed to describe this heterogeneous pattern of early Phanerozoic diversification. The model adequately describes the “multiple equilibria,” the asymmetrical history of the “Cambrian fauna,” the extremely slow initial diversification of the later “Paleozoic fauna,” and the combined patterns of origination and extinction in both faunas. It is suggested that this entire pattern of diversification reflects the early success of ecologically generalized taxa and their later replacement by more specialized taxa.
New perspectives in vertebrate paleoecology from a recent bone assemblage
- Anna K. Behrensmeyer, David Western, Dorothy E. Dechant Boaz
-
- Published online by Cambridge University Press:
- 08 February 2016, pp. 12-21
-
- Article
- Export citation
-
Interpretations of vertebrate paleoecology depend on knowledge of taphonomical processes which alter the composition of the preserved fossil assemblage from that of the original community. Study of the potential fossil record of a recent mammal community in Amboseli National Park, Kenya, shows the effects of some of these biasing processes and demonstrates how a bone assemblage on a modern land surface can be a source of past and present ecological information. In the bone assemblage, species presence or absence and relative abundance differ from recorded living species occurrences and population sizes: only 74% of the extant species in the basin are identified in the bone sample, and carcass abundances vary significantly from known population sizes of the major herbivore species. Both biases appear to be strongly correlated to body size, and this results from greater destruction of bones of smaller animals within the weight range from about 1-1000 kg. This size-biasing against small species appears to be due primarily to the greater susceptibility of small bones to destruction by carnivore mastication, breakage through bioturbation (trampling), and physical and chemical processes of weathering. Size-biasing resulting from such primary processes can thus be inherited by buried bone assemblages whatever their final mode of deposition. The bone assemblage also provides information on the spatial distributions of the major herbivore species over six major habitats. Patterns of strong habitat specificity are accurately represented in the bone assemblage. However, the record for certain species is affected by their seasonal and diurnal habitat shifts so that their bone distributions do not match live census data. The Amboseli bone assemblage provides a modern analogue for taphonomical processes which may have affected fossil assemblages derived from paleo-land surfaces prior to fluvial transport. It also helps to define limits of resolution in interpreting paleoecological information from such fossil assemblages.
Paleobiological and isotopic studies of eggshells from a declining dinosaur species
- H. K. Erben, J. Hoefs, K. H. Wedepohl
-
- Published online by Cambridge University Press:
- 08 February 2016, pp. 380-414
-
- Article
- Export citation
-
Late Cretaceous dinosaur eggshells from southern France and the Spanish Pyrenees, presumably belonging to the sauropod Hypselosaurus priscus Matheron, are almost exclusively composed of primary calcite. Besides normal development of these eggshells, there appear two kinds of pathologic tendencies: bi- or multi-shells (infrequent), and shells with a reduced thickness (increasing in frequency until, in the uppermost horizon, they represent more than 90% of the sample). The extinction of the species is attributed primarily to the consequences of thinning of the eggshells.
The physiological mechanisms producing pathologic dinosaur eggshells are evaluated in the light of homologous phenomena occurring in living birds and reptiles. On this basis, it is concluded that in the late Maastrichtian populations of “Hypselosaurus,” pathologic eggshells were caused by hormonal imbalances of the vasotocin and of the estrogen levels. On the same basis it is postulated that the teratological shell repetition led to embryo suffocation and that the pathological reduction in shell thickness caused shell breakage and dehydration of the embryo. The lethal results are evident from the frequent absence of “resorption craters” in the mammillary knobs of pathologic shells, a fact which indicates either lack of fertilization of the eggs or the perishing of the embryo prior to the calcification of its skeletal bones. A change in environmental conditions is the ultimate factor which caused hormonal imbalances and extinction. Such a change is indicated by a shift of the mean oxygen isotopic composition (δ18O) of eggshell carbonates from −0.6%o to −5.3%o, and by changes in Sr. Information of palaeo-climate is primarily derived from eggshells of living birds and reptiles. The correlation between temperature and oxygen isotopic composition of waters (and related carbonates) is less distinct than for marine carbonates. δ13C ranges from −16.5 to −4.5 of eggshells of extant species indicate food from “normal” C3 metabolism and from C4 metabolism of plants in a dry climate. “Hypselosaurus” populations probably consumed “normal” C3 plants. Using isotopic calibration of eggshell carbonates for the interpretation of δ13C and δ18O values of dinosaur eggshells, a slight change from higher to lower temperatures or a change from a dry to a more humid climate during the time from Lower (and Middle) to Late Maastrichtian can be assumed. The latter explanation is favored because the exceptionally high Sr in the Early Maastrichtian eggshells could be a potential indicator of co-existing evaporites.
Maximum likelihood estimation of fossil assemblage composition
- Richard C. Holtzman
-
- Published online by Cambridge University Press:
- 08 February 2016, pp. 77-89
-
- Article
- Export citation
-
The most common methods of estimating the relative abundance of species in a fossil assemblage are all maximum likelihood estimates. They differ from one another in their inherent assumptions made about the effects of fragmentation and differential preservation in the assemblage. For many fossil assemblages, relative abundance is best estimated by the relative frequency of specimens or relative frequency of elements. Monte Carlo simulations suggest, however, that in most other circumstances estimates based on frequency of elements divided by the number of elements in a complete individual provide greater accuracy than estimates based on minimum number of individuals. This relation results from an interaction between random sampling error and a variety of biases inherent in the two estimates.
Apparent prolonged evolutionary stasis in the middle Eocene hoofed mammal Hyopsodus
- Robert M. West
-
- Published online by Cambridge University Press:
- 08 February 2016, pp. 252-260
-
- Article
- Export citation
-
The stratophenetic approach to phylogeny reconstruction, developed in numerous recent papers by Gingerich, has its strongest data base in comprehensive collections from early Eocene rocks in the Big Horn Basin of northwestern Wyoming. Other examples have been criticised for lack of well-documented collection data, making suitable testing of stratophenetic interpretations of evolutionary modes virtually impossible.
Study of Hyopsodus from the middle Eocene Bridger Formation of southwestern Wyoming provides new collection data. Adequate stratigraphic documentation is available, so Gingerich's graphical presentation can be essentially duplicated. In contrast to the complex branching pattern Gingerich interpreted from his data, Bridger Formation Hyopsodus data seems to show little size change through approximately one million years. This stasis or equilibrium condition, which is considered rare by Bookstein et al. (1978), is the only well developed pattern in Bridger Hyopsodus.
Preserved ligament in a radiolitid rudist bivalve and its implication of mantle marginal feeding in the group
- Peter W. Skelton
-
- Published online by Cambridge University Press:
- 08 February 2016, pp. 90-106
-
- Article
- Export citation
-
Fibrous ligament and aragonitic shell material have been found preserved in two specimens of radiolitid rudists [cf. Radiolites angeoides (De Lapeirouse)] from the Lower to Middle Coniacian of Tirol, Austria. The invaginated ligament apparently comprised two torn and overgrown non-functional strings, one in each valve, connected medially by a thin strip of active ligament. The original dimensions of the active ligament have been estimated, for shells of 4 to 5 cm commissural diameter, at 0.58 mm2 cross-sectional area and 1.3 mm length, with the shell closed. Using data on the adductor muscle of Recent oysters and the results of compression tests carried out on samples of oyster fibrous ligament, it is estimated that these radiolitid shells opened less than 1 mm in life; their gaping was highly restricted. This is taken to support the proposal that the expanded mantle margins (principally that of the right, attached valve) were responsible for food particle entrapment, rather than the gills as in most other bivalves. It is suggested that a tentaculate mantle flange was extruded through the narrow gap between the valves during feeding. The pair of pronounced radial bands on the posterior flanks of the shell are interpreted as marking pseudofaecal and faecal ejection sites. Finally, it is speculated that slime (possibly poisonous or foul-tasting) on the mantle tentacles inhibited predators.
Functional morphology of Cretaceous helically-coiled ammonite shells
- Peter Ward
-
- Published online by Cambridge University Press:
- 08 February 2016, pp. 415-422
-
- Article
- Export citation
-
Torticonic (helically-coiled) ammonoids have been most commonly interpreted as vagile, benthonic forms. Their mode of coiling, however, places the siphuncle in a functionally dorsolateral, rather than ventral position on the shell whorl due to: 1) their probable apex-upward shell orientation during life, and 2) size asymmetry of the lateral sutural saddles due to helical coiling, associated with upward siphuncle displacement. The resultant positioning of the siphuncle produces cameral liquid de-coupling soon after initiation of cameral liquid removal (emptying of liquid from a newly formed chamber within the phragmocone). Since in Recent chambered cephalopods cameral liquid de-coupling increases efficiency of vertical migration, a similar mode of life was indicated for the torticonic ammonites.
The environment of Ramapithecus in Africa
- Peter Andrews, Elizabeth Nesbit Evans
-
- Published online by Cambridge University Press:
- 08 February 2016, pp. 22-30
-
- Article
- Export citation
-
The faunas of three fossil sites in Africa that contain time successive groups of fossil Hominoidea have been analyzed to determine their paleoenvironments. The basis for the analysis is an assessment of the ecological diversity of the fauna, which is expressed in terms of four categories: taxonomic composition, body size, feeding habits and locomotor zonal adaptation. This method has shown that the community structures of the three fossil faunas are significantly different, and comparisons with the community structure of modern habitats suggest that the environment of the early Miocene fauna of Songhor, and the primitive apes associated with it, was probably a type of lowland forest; the habitat of Ramapithecus and the Fort Ternan middle Miocene fauna compares best with modern woodland-bushland habitats; and the habitat of Homo habilis at Olduvai appears to have been intermediate between grassland and woodland-bushland. If man evolved from one of the early forest living apes, as seems likely on present evidence, an adaptive shift from forest to non-forest habitats must have occurred at some stage in his evolution. The evidence from Fort Ternan shows that in Africa Ramapithecus made this adaptive shift, and it is also now becoming clear that several genera of Eurasian apes, including Ramapithecus, made a similar environmental change at the same time.
Disarticulation and scattering of mammal skeletons
- Andrew Hill
-
- Published online by Cambridge University Press:
- 08 February 2016, pp. 261-274
-
- Article
- Export citation
-
I present a statistical technique for determining the disarticulation sequence of vertebrate skeletons based on the relative numbers of different intact joints in an assemblage of bones. For remains of modern Topi (Damaliscus korrigum) on the margin of Lake Turkana, northern Kenya, the disarticulation pattern is very consistent. This sequence, on dry land, differs from that reported for bovids that have disarticulated in the presence of water. On land the bones first released as single bones are those moved least easily by currents of moving water. The last released are those moved most easily. I develop a model of random scattering that suggests that the rate of dispersion is great at high concentrations of bones and decreases rapidly as the distance between bones increases. This leads to a condition where scattering effectively stops. The area of more or less stabilised dispersion is dependent only upon the mean distance that each random event moves a bone. Tests show that it is unlikely that articulated units themselves are much involved in scattering, and scattering appears to take place throughout the course of disarticulation.
Mode of life of planktonic graptolites: flotation structure in Ordovician Dicellograptus sp.
- Stanley C. Finney
-
- Published online by Cambridge University Press:
- 08 February 2016, pp. 31-39
-
- Article
- Export citation
-
An unusual isolated specimen representing an immature rhabdosome of Dicellograptus sp. from the Middle Ordovician Athens Shale of Alabama has a flattened sphere attached to the distal end of the nema. The flattened sphere can be interpreted as a float that probably conferred buoyancy on the rhabdosome. Such floats, which became detached in mature rhabdosomes, may be common in the fossil record. The presence of nematophorous siculae in all planktonic graptolites suggests that for immature rhabdosomes a passive, buoyant mode of life, attained by means of a float, might have been widespread taxonomically. This mode of life is compatible with the theory of passive response for the mode of life of mature rhabdosomes, but it is difficult to reconcile with the theory of automobility.
Rarefaction and rarefiction—the use and abuse of a method in paleoecology
- John C. Tipper
-
- Published online by Cambridge University Press:
- 08 February 2016, pp. 423-434
-
- Article
- Export citation
-
Rarefaction is a method for comparing community diversities that has consistently been abused by paleoecologists: here its assumptions are clarified and advice given on its application. Rarefaction should be restricted to comparison of collections from communities that are taxonomically similar and from similar habitats: the collections should have been obtained by using standardised procedures. The rarefaction curve is a graph of the estimated species richness of sub-samples drawn from a collection, plotted against the size of sub-sample: it is a deterministic transform of the collection's species-abundance distribution. Although rarefaction curves can be compared statistically, it may be more efficient to compare the species-abundance distributions directly. Both types of comparison are discussed in detail.
Deep-sea foraging pathways: an analysis of randomness and resource exploitation
- Jennifer A. Kitchell
-
- Published online by Cambridge University Press:
- 08 February 2016, pp. 107-125
-
- Article
- Export citation
-
The foraging paradigm of trace fossil theory has historically accorded random behavior to non-food-limited deposit-feeders and non-random behavior to food-limited feeders. A series of randomness measures derived from empirical modeling, simulation modeling, stochastic modeling and probability theory applied to foraging patterns observed in deep-sea bottom photographs from the Arctic and Antarctic yielded a behavioral continuum of increasing non-randomness. A linear regression of trace positions along the continuum to bathymetric data did not substantiate the optimal foraging efficiency-depth dependence model of trace fossil theory, except that all traces exhibited a greater optimization than that of simulated random foraging. It is hypothesized that optimization as evidenced by non-random foraging strategies represents maximization of the cost/benefit ratio of resource exploitation to risk of predation and that individual foraging patterns reflect an exploration response to the morphometry of a patchily distributed food resource. Differential predation and competition may account for the co-occurrence of random and non-random strategies within the same bathymetric zone.
Cameral liquid in Nautilus and ammonites
- Peter D. Ward
-
- Published online by Cambridge University Press:
- 08 February 2016, pp. 40-49
-
- Article
- Export citation
-
Recent Nautilus pompilius from the Fiji Islands and N. macromaphalus from New Caledonia show decreasing cameral liquid volumes relative to total phragmocone volume during ontogeny. A maximal value of 32% of the phragmocone filled with cameral liquid was measured from a 190 g N. pompilius. No specimens of over 500 g total weight of either species exceeded 12%. These figures are in contrast to values derived for seven ammonoid species by Heptonstall (1970), who found values ranging between 19 and 52%.
The relationship between cameral liquid volume and salinity within single chambers engaged in the emptying process are examined in N. pompilius and N. macromphalus. Both species start with newly formed chambers filled with cameral liquid isotonic to seawater. Ionic removal by the siphuncular epithelium rapidly reduces the cameral liquid osmolarity, producing osmotic movement of the cameral liquid into the blood spaces of the siphuncle. In both species the lowest cameral liquid salinities occur when the chamber is slightly over half emptied. After this point, which coincides with decoupling of the cameral liquid from the siphuncle, cameral liquid volume continues to decrease, but cameral liquid salinity increases, indicating that the rate of ionic removal slows relative to liquid removal. In N. macromphalus decoupled cameral liquid salinity rises until it is nearly isotonic to seawater when the chamber is nearly emptied. In N. pompilius, however, the rate of ion removal in decoupled cameral liquid is not slowed as much as in N. macromphalus, since it rarely exceeds 40% seawater osmolarity even when the chamber is nearly emptied. The differences in emptying methods demonstrated in these two species are probably related to their different habitat depths: N. pompilius from Fiji is found in much deeper water and must employ more physiologic work to empty chambers at greater depth.
A model for paleobiogeography of South American cricetine rodents
- Larry G. Marshall
-
- Published online by Cambridge University Press:
- 08 February 2016, pp. 126-132
-
- Article
- Export citation
-
A model for the paleobiogeographic history of South American cricetine rodents is proposed based on new and/or recently published fossil, geological, paleobotanical and radioisotope data. Cricetine rodents of the tribe Sigmodontini evolved in North America before 7.0 Myr BP. They got to South America by waif dispersal across the Bolivar Trough marine barrier from Central America during a world wide drop in sea level (the “Messinian Low”) between 7.0 and 5.0 Myr BP. The basal stock was probably a sylvan (forest) form, from which evolved pastoral (grazing) forms in the savanna-grassland area of Venezuela, Colombia and the Guianas. The pastoral forms in the northern savanna-grassland area were restricted there until about 3.5 Myr BP. At that time there occurred the first glaciation in South America and consonant with glacial advance was a retraction of forest habitats and an expansion of savanna-grassland habitats. At that time the pastoral forms were able to disperse southward through a savanna-grassland corridor along the eastern foothills of the Andes and spread throughout the previously disjunct savanna-grasslands of Bolivia and Argentina. Cricetines are first recorded as fossil in the Monte Hermoso Fm. of Argentina which is about 3.5 Myr BP in age. The Panamanian land bridge came into existence about 3.0 Myr BP as indicated by the beginning of a major interchange of terrestrial faunas between the Americas, which was well underway by 2.7 Myr BP.
Taphonomic investigations of owl pellets
- Peter Dodson, Diane Wexlar
-
- Published online by Cambridge University Press:
- 08 February 2016, pp. 275-284
-
- Article
- Export citation
-
Owls are important consumers of small vertebrates, and because they regurgitate pellets rich in bone, they may be important potential contributors of the concentrated remains of small vertebrates to the fossil record. Owls of three sizes, the large great horned owl (Bubo virginianus), the medium-sized barn owl (Tyto alba), and the small screech owl (Otus asio), were fed a common diet of mice. The bony contents of the pellets were analyzed to determine the amount of bone loss by digestion, bone completeness, and sites of bone breakage. For all three species, only about half the number of bones ingested were recovered in the pellets. Mandibles and femora were most abundant, and pelves and scapulae were the least abundant. Screech owls broke 80% of the cranial and limb elements, barn owls only 30%. Skulls fared poorly in great horned and screech owl pellets, while barn owls returned 80% of the skulls intact, with only the caudal portion of the cranium damaged; barn owls also returned articulated strings of vertebrae and complete paws. These results provide a baseline for the recognition of owls as agents of accumulation of small bones in the fossil record and suggest that the actions of ancient predators may be revealed by species-specific patterns of bone destruction of an assemblage of fossil prey species.
Interpretation of paleoecological similarity matrices
- E. C. Pielou
-
- Published online by Cambridge University Press:
- 08 February 2016, pp. 435-443
-
- Article
- Export citation
-
Consider a one-dimensional sequence (temporal or spatial) of samples from a many-species community. The sequence may exhibit no steady change, or gradual, steady change, or one or more abrupt, stepwise changes. The pattern is graphically displayed in a similarity matrix, whose elements are the pairwise similarities between every possible pair of samples.
Statistical study of the patterns of similarity matrices will be shown to be informative. A criterion, Q, is proposed for measuring the “disarray” of a similarity matrix; matrices constructed from “perfect” sequences, exhibiting uninterrupted steady change, have Q = 0. The probability distribution of Q in “random” similarity matrices (constructed from sequences in which the ordering is wholly haphazard) is examined.
Several applications are illustrated, using data on foraminiferal assemblages in marine sediment cores as examples. Use of the criterion permits the following: (1) Objective recognition of regions in which stratigraphy is especially well preserved. (2) Selection for study of cores in which the sequence of change is especially clear, and rejection of cores in which the natural sequence has been obscured by reworking. (3) Judgment as to whether, in sequences exhibiting stepwise changes, gradual change between the steps occurs as well. (4) Discrimination between species-sets that track gradual environmental change and sets that switch abruptly. (5) Objective choice of the best similarity index for measuring pairwise similarities in a given body of data.