304 index pageZOOL 304, Class Notes
Chapter 2, Jan. 20 - Jan. 24.
Text reading: Chapter 2.
- Assignment for written preparation.
- Discussion.
- Introduction
- What is adaptation? (definitions)
- How can adaptation be demonstrated?
- Comment on text Chapter 2.
- Characteristics of adaptive evolution
- Examples of adaptive evolution
- Types of selection
- Concluding sections
- Additional page on the analysis of adaptation
- Study questions for text reading.
Notes for chapter 1 / 2 / 3 / 4 / 5 / 6 / 7 / 8 / 9 / 10 / 11 / 12 / 13 / 14 / 15 / 16 / 17
304 index pageIntroduction
Living creatures are wonderfully well suited to survive and reproduce in so many diverse circumstances, including many that seem quite hostile to life. Living creatures also display a multitude of amazing and often puzzling traits of form and behavior. Biologists have long sought answers to questions of two related types.
- How do organisms work? That is, what particular "adaptations" enable each species to live in its particular niche.
- What is each trait for? That is, what adaptive role does each outstanding trait (body part or behavior) play in survival and reproduction?
Such questions may be asked and answered as if each organism had been "designed" to fit its current niche. Although this approach is central to certain biological disciplines, notably physiology, it is incomplete until it also addresses the essential questions of how each particular trait and role have evolved. In much biology, only lip-service is played to evolutionary explanations of adapatation, in which past processes of variation and selection are simply presumed rather than exposed to critical investigation.
Here we begin to examine adaptation from an evolutionary perspective.
Chapter 1 (p. 12) emphasized the relationship between adaptive evolution and natural selection. In the usage of our authors, natural selection refers to "the correlation of reproductive success with a trait." Adaptive evolution refers to "the complete process, including the genetic response to selection" (i.e., the change in allele frequencies which results from the correlation of hereditary variation with variation in reproductive success).
Chapter 1 (p. 13) also points out that adaptation refers to both a process and a condition. And this chapter notes significantly that "adaptation can be hard to demonstrate" [as a true evolutionary process involving variation and selection]. Also offered were some examples of conditions that suggest adaptation and a list of circumstances which may set limits to adaptation.
Chapter 2 introduces several major topics associated with the concept of adaptative evolution:
- The "subtlety and power" of natural selection.
- The diverse causes of natural selection.
- Examples of natural selection.
- Evidence of natural selection in action.
- Types of natural selection.
IMPORTANT NOTE: Proper consideration of causation implies that descriptions which include phrases like "adapted for" or "adaptation for" (e.g., "eyes are adapted for vision", "claws are an adaption for prey capture", etc.) are potentially misleading. The process of adaptive evolution does not build for any current or future function. It only accumulates variants which have worked successfully in the past.
This "adapted for" usage is convenient shorthand and is deeply ingrained in our use of language. Every student of evolution is therefore cautioned to think clearly about the implications of this phrase. "Trait X is adapted for function Y" means "Selection has acted in the past through a correlation of variation in trait X with variation in reproductive success AND the correlation was causally mediated by function Y." To demonstrate that a trait is an adaptation for a function, it is necessary to demonstrate that this evolutionary process has indeed taken place. And such demonstration is seldom simple.
Definitions
Before continuing, it is worth noting that usage of the word "adaptation" is somewhat problematic. The word as commonly used has several different meanings, some of which are unrelated to evolution. Even when used in an evolutionary sense, "adaptation" is used with some subtle variations in meaning that can lead to misunderstanding. Thus we continue with a discussion of the word itself.
Words such as adaptation, adapt, adaptedness, and adaptive are derived from the Latin ad + the root aptus = "fitted, suited, appropriate"; see OED definition.
In common, everyday usage, these words all refer to the suitability (aptitude) of organisms for successful life in particular niches. However, this usage conflates several different causes for such suitability.
Developmental adaptation is something that individual organisms can do during their individual lifetimes. Similarly, physiological adaptation is something that individual organisms can do to adjust to immediate needs. In contrast, evolutionary adaptation is something that only populations can do, over many generations during which heritable variation is selected by the conditions of existence which establish a correlation between heritable variation and variation in reproductive success.
Since all of these usages are very common in biology, they must be carefully distinguished. This is often neglected in popular writing, leading to serious misunderstanding -- such as a belief that evolutionary adaptation follows from physiological adaptation (the idea of inheritance of acquired characteristics) or a belief that the process of adaptation is directed toward the acquisition of a useful trait.
Unfortunately, determining whether some particular trait is based on physiological adaptation or evolutionary adaptation may be difficult and can require arduous and carefully controlled experimentation.
As a potential source of some additional confusion, the word "adaptation may refer to:
- a process of adapting,
- a condition of adaptedness, or
- a trait called an adaptation.
Thus adaptation (the process) leads to adaptation (the condition) through acquisition of an adaptation (the trait). An individual organism becomes physiologically adapted through internally controlled modification of its individual circumstances. A population becomes evolutionarily adapted through selection of heritable traits over many generations.
When used in a strong evolutionary sense, adaptation should ideally refer to causation by natural selection acting on heritable variation. According to this usage, traits should be called adaptations only if they have been shaped by the process of natural selection acting on heritable variation. This process may be called "adaptive evolution", or just "adaptation".
However, since we seldom have thorough evidence of the history of causation for any trait, assertions of adaptation are usually best understood as hypotheses (proposed with varying degrees of confidence) rather than as well-tested conclusions.
Also see IMPORTANT NOTE above.
How do we know whether or not some trait is an adaptation?
There are at least two basic approaches.
- Presume that that any given trait is likely to be adaptive, especially if it exhibits apparent functionality and/or complexity.
- Begin with skeptism about the adaptiveness of any given trait.
Presume that a trait is NOT adaptive UNLESS it passes certain stringent tests.The first approach is sometimes characterized as adaptationism. Although the label "adaptationist" is typically intended perjoratively (implying uncritical acceptance of an untested hypothesis of adaptation), this approach is standard operating procedure in several domains of biology (notably physiology).
- Commonly accepted criteria for presuming adaption include:
- obvious contribution to fitness (compared to a putative ancestral condition).
- functional complexity.
Unfortunately, this approach encounters several difficulties.
- The criteria do not readily distinguish between cause and effect.
Some traits may indeed be useful for some function without ever having been shaped by selection (more below). Some former adaptations may be "obsolete"; that is, traits which were shaped by selection at some time in the past, may no longer contribute to fitness. And many useful (even essential) traits originated under different circumstances and are perhaps best understood not as adaptations to current conditions but as inheritance from distant ancestors.
- Some traits appear to have originated earlier in time than the functions which they are presumed to serve.
This is the problem of "preadaptation", which seems to violate basic principles of evolutionary causation.
- Experience has taught that plausible explanations for how a trait contributes to fitness are too easily invented.
Overly facile adaptive rationales are sometimes called "just-so stories" (after Rudyard Kipling's fables -- how the elephant got its trunk, how the camel got its hump, etc.). The tendency to invent plausible explanations has been caricatured as the "Panglossian paradigm" (after fictional character in Voltaire's novel Candide; the philosoper Dr. Pangloss professed that everything had a purpose and that all events were always for the best).
See Gould, S.J. and R.C.Lewontin (1979), The spandrels of San Marco and the Panglossian paradigm: A critique of the adaptationist programme. Proc. Royal Soc. London 205:281-288.
Skepticism is a good antidote to Panglossian excesses. On the other hand, too much skepticism can interfere with progress. The functions for many traits are far from obvious. Without a prior presumption of adaptive value as motivation for further study, such functions might never be discovered. Furthermore, few tests for adaptation yield unambiguous results.
Tests for adaptation.
Generally, adaptation may be reasonably inferred (but not proven) whenever there is a plausible proposal for functionality (i.e., for a causal relationship between the putatively adaptive trait and reproductive success, such that prior variants would be expected to have lesser fitness) that is countered NEITHER by any indication that the trait has always been invariant NOR by any indication that the trait was present prior to its putative adaptive role.
Such analysis is always complicated by the fact that evolution is a process where history matters.
- Traits which were adaptive in the past may no longer contribute to fitness. (Such traits might more accurately be called "obsolete" adaptations.)
- A trait which currently contributes to fitness may not have been shaped by selection for its current function, so that its currently function is fortuitous rather than properly adaptive. At least three different modes for such origin are understood. A new term, "exaptation", was coined as a label for a trait has a useful function by virtue of its form but was not selected for this function (Gould, S.J., & Vrba, E.S. 1982. Exaptation: A missing term in the science of form. Paleobiology 8: 415).
- Traits may have been originally shaped by selection for some entirely different function. (It is helpful if proposals of adaptation specify when and on what basis the trait has been selected.)
- Traits which contribute to fitness may originate as inevitable effects of basic physical/chemical principles. For example, everything must be some color. So any particular color -- even a spectacular color like bright red blood -- may not need an adaptive explanation.
- Traits which contribute to fitness may emerge as incidental effects of genetic and developmental programs which evolved for other reasons.
- The same constraints which can limit adaptation (see chapter 1, pp. 15-18, 25-29) may also maintain an appearance of adaptation when selectable variation is not available.
Adaptation is also suggested by a measurable correlation of heritable trait variation with variation in reproductive success, so that stabilizing or directional selection appears to be maintaining the trait. However, the standard caveat applies: Correlation is not causation.
A trait may correlate with fitness not because that trait itself contributes to fitness but because it is linked (by genetic linkage or by pleiotropy) with some other trait that has an effect on fitness. (That is, a test which demonstrates that evolution is occurring does not necessarily indicate reliably what trait is responsible for conferring the selective advantage.)
Also, traits that do not vary will display no correlation with fitness, but absence of such correlation does not disprove adaptation. An adaptive trait may have been shaped by selection in the past and then become so thoroughly canalized that it is no longer independently affected by genetic variation.
Another test for adaptation is correlation between trait and niche among a range of populations or species. But again, results may be ambiguous. Interpretation may be confounded if the correlation is mediated by developmental or physiological response of the trait to the niche environment or if both trait and niche are correlated with ancestry. These confounding hypotheses must be ruled by additional tests
The possibility of physiological or developmental adaptation is typically controlled for by raising individuals from the same population species in different environments.
The possibility that a correlation between trait and niche may be due to ancestry rather than to adaptation is typically controlled by phylogenetic analysis. This requires BOTH that the phylogeny be known or plausibly inferred AND that this phylogeny includes several separate branches that each include variation in both trait state and niche.
Various other tests can also lend indirect support to a hypothesis of adaptation, but true evolutionary adaptation is notoriously difficult to prove. Unfortunately, well-controlled tests for adaptation are often arduous if not impossible.
These difficulties suggest that adaptation may in many cases be reasonably inferred, but that any such inference should be accompanied by due consideration of alternative hypotheses. And the conclusion may seldom be a confident assertion of fact but rather a cautious hypothesis of plausibility.
304 index page
304 index pageChapter 2 is divided into several sections.
- Characteristics of adaptive evolution
- Examples of adaptive evolution
- Types of selection
- Concluding sections
The introductory section (p. 35) describes some fundamental characteristics of adaptive evolution.
- Evolution is noted for "the subtlety and power of natural selection".
- Evolution's causes seem ordinary, but its consequences seem extraordinary.
- Natural selection acts at several levels.
- Natural selection acts on phenotypes but effects genotypes.
- Also see concluding sections below.
The next section (pp. 36-46) describes several examples of natural selection in action. Here the important learning objective is to understand what each example illustrates about the process of natural selection.
- Galapagos finches: Variation in reproductive success can be considerable, leading to very strong selection. (For more about the finches, also see p. 88, Chapter 4. For a bit of background, click here. For lots more, read The Beak of the Finch: A Story of Evolution in Our Time, by Jonathan Weiner.)
- Lotus corniculatus: The costs of a particular adaptation can lead to selection for polymorphism. (Incidently, as suggested in the accompanying text-drawing, Lotus corniculatus is a legume, family Fabacae, also called bird's-foot-trefoil [for image and description, click here, then cut-and-paste the name into the search function].)
- Kestrels: Reproductive success is more subtle than simply producing the most offspring at every opportunity. Maximal lifetime reproductive performance may call for a compromise between current reproduction and future reproductive opportunity. Try to understand the significant idea of residual reproductive value (p. 39), which is a statistical expectation of future potential.
- Flowers and pollinators: Coevolution can result in remarkable correspondence among interacting species.
- Mussels: Allele frequencies which are not in equilibrium suggest that selection must be occurring.
- Antibiotic resistance: Adaptive evolution can be rapid and effective, especially in large populations with extensive genetic variation and short generation time subjected to strong selection.
The next section (pp. 45-51) describes the vocabulary commonly used in evolutionary biology to classify the general types of selection.
- According to effects on traits which vary quantitatively:
- Stabilizing selection (greater reproductive success for individuals closer to the population average)
- Directional selection (greater reproductive success for individuals on one side of the population average)
- Disruptive selection (lower reproductive success of individuals closer to the population average)
- According to whether or not trait variation affects mating success:
- Sexual selection
- Natural selection (by convention, any selection which is NOT based on mating success)
- According to ecological circumstances. The text discussion of density-dependence is sketchy and inadequate. However, the basic logic that underlies density dependence is discussed in Chapter 8, especially pp. 156-160.
- Density-dependent selection commonly affects traits which directly impact the maximal rate of population growth -- traits such as number of offspring, size of offspring, and time to maturation. Traits which promote rapid population growth are selected when population size is far below carrying capacity, when the best way to have more surviving offspring is simply to produce more offspring. (This circumstance is sometimes called "r-selection", where r stands for rate of population growth.) Traits which promote survival of offspring in competition for limited resources are selected when population size is close to carrying capacity, when the best way to have more surviving offspring is to assure that each offspring produced has the best chance of success. (This circumstance is sometimes called "K-selection", where K stands for carrying capacity.)
- Density-independent describes selection acting on traits whose contribution to fitness does not depend on how close the population is to its carrying capacity. Density-independent selection affects most traits.
- According to frequency within a population. For more on the consequences of frequency-dependent selection, see pp. 105-107 in chapter 5.
- Frequency-dependent describes selection in which the relative fitness of a trait (or allele) depends on whether that trait (or allele) is common or rare in a population. Examples include selection based on the hunting patterns of predators or parasites (e.g., the side-biting, scale-eating fish, text chapter 5, pp. 105-106 [Hori, M. 1993, Frequency-dependent natural selection in the handedness of scale-eating fish. Science 260:216-219]) and strategies for gaining access to mates (e.g., the rock-paper-scissors lizards, more; Smith, John Maynard. "The games lizards play." Nature 380:198-199). For more examples, just cut-and-paste the phrase "frequency dependent selection" into an internet search engine such as Yahoo.
- Frequency-independent describes any selection which is NOT frequency-dependent.
- According to the "level" or "unit" of biological organization upon which selection operates.
- Individual selection. Unless there is compelling reason to believe otherwise, most selection is presumed to act on individual organisms
- Kin selection. Selection in which traits are favored because they increase the reproductive success of close relatives ("kin") who are likely to carry corresponding alleles.
- Group selection. Selection based on traits which are attributes of groups rather than individuals; this mode of selection has been difficult to support, since most suggested "group-traits" seem to be explainable on the basis of individual selection or kin selection.
- Species selection. Selection based on differential rates of speciation and/or extinction. Logically plausible but controversial, since no good examples have been well-demonstrated. The time course of speciation is so long compared to the generation time of individuals, that individual selection ought to be a much faster, more effective process. In other words, there are NO good examples of traits which have evolved "for the good of the species."
- According to the relationship between selected genotype and selected phenotype.
- Direct selection is based on a direct, causal relationship between the phenotypic affects of the selected gene and relative reproductive success. We normally presume that selection is actingly directly.
- Indirect selection is, well, indirect. The best-understood example is genetic "hitchhiking", in which genetic variation correlates with reproductive success simply because of genetic linkage to a directly-selected gene.
The next section (pp. 51-52) briefly discusses the strength of selection (it may be weak or strong) and the rate of evolutionary response (it may be remarkably rapid or extremely slow; it may be gradual or episodic). For examples, see pp. 2-3 (South American guppies), pp. 36-37 & 88 (Galapagos finches), and pp. 42-44 (antibiotic resistance).
The next short section (pp. 52-53) reminds us that selection is always context dependent. That is, the relative success of particular traits depends on the ecological circumstances in which selection occurs and on the genetic background against which those traits appear.
It should also be emphasized here that adaptation, like fitness, is relative; its assessment depends on comparison among different variants. As always, evolution is all about variation. When we refer to a feature simply as "an adaptation ", the implied comparison is with a prior, ancestral variant of the trait.
For example, a statement that "Thick fur is an adaptation" implies that "At some time in the past, ancestors having thick fur reproduced more successfully than others in the same population which had less-thick fur." To refer to a complex trait as an adaptation (e.g., "Eyes are an adaptation"), one must presume a long history during which a particular ancestor-to-descendent sequence of variants reproduced more successfully than alternative variants.
The final section (p. 53) reminds us that the basic logic of adaptive evolution based on relative reproductive success correlated with heritable variation may apply in domains other than biological evolution, domains such as human culture and computer algorithms.
Notes for chapter 1 / 2 / 3 / 4 / 5 / 6 / 7 / 8 / 9 / 10 / 11 / 12 / 13 / 14 / 15 / 16 / 17
Comments and questions: dgking@siu.edu
Department of Zoology e-mail: zoology@zoology.siu.edu
Comments and questions related to web server: webmaster@science.siu.edu