Schmalhausen and the Baldwin Effect

With the aim of finding a place for somatic adaptation in evolution, James Mark Baldwin (1861-1934) and others proposed the hypothesis of organic selection, which drew on Lamarck and Darwin without conflating them. Baldwin was an early experimental psychologist at a time when psychology was separating from philosophy as a field of study. Behavior had all the elements of somatic adaptation, and Baldwin proposed that animals have broad ranges of somatic adaptability enabling them to tolerate environmental change when their niche alters or when they enter new niches.

In the unfamiliar environment, the organism is stressed; hence it continues to utilize its adaptive mechanisms. It is not fully adapted, though viable enough to reproduce at least minimally. In subsequent generations, heritable changes arise in a few members ofthe population that fix the somatic adaptation and remove stress; these members are then selected for their increased reproductive fitness. In this way a heritable internal stimulus has replaced the external one in maintaining the adaptation. Thereafter the organism expresses it even when removed from the environmental stimulus.3

This hypothesis differs in several respects from the neo-Darwinian view, although to make the comparison we must translate some of Baldwin's pre-Mendelian language into modern terms. According to Baldwin, mutant variants would not have to precede selection: instead, mutation could follow selection. In the new condition, somatic adaptation would suffice at first for the survival of many members of the population, not just rare variants. Mutants and in particular new gene combinations would in time arise in a somatically adapted population. They would be selected as they stabilized, refined, and extended the somatic adaptation.

This phase of the process is simply Darwinian variation and selection. The somatic adaptation becomes a heritable evolutionary adaptation, persisting even under nonstress conditions when these arise. For Baldwin, mutation might do more than stabilize the somatic adaptation; it could stabilize it at a more optimal state.

Baldwin's proposal did not violate the Darwinian variation-selection principles, yet it gave prominence to individual somatic adaptability. Without that preexisting adaptability, the new selective conditions might extinguish the organism before it had a chance to adapt genetically. Baldwin avoided Lamarck's inheritance of acquired characteristics, which were somatic adaptations, because Baldwin required independent heritable genetic change. That took the form of new genetic combinations from the genetic variability in the population or new mutations to stabilize and fully express the adaptation. The main implication is that the complex phenotypic variation is not created from nothing, but rather from preexisting processes and components of the organism's somatically adaptable phenotype, whereas mutations merely stabilize and extend what is already there. Thus, the mutations need not be creative and numerous, a proposal that of course greatly reduces the difficulty in generating phenotypic change.

Ivan Ivanovich Schmalhausen (1884-1963) expanded the Baldwin proposals in his book Factors of Evolution, completed in Moscow in 1943 in the depths of World War II while the city was under Nazi siege. Schmalhausen was a leading figure in Russian biology before the war and until 1948. At that point Fyodor Lysenko, the notorious Stalinist scourge of Russian genetics, brought him to trial for being a "Weismannist-Mendelist-Morganist idealist."4

Schmalhausen began with the concept of the "norm of reaction" of an organism, that is, the range of phenotypes expressed when the organism reacts to various environmental conditions (temperature, humidity, crowding, kind of food). This reaction norm has two components. When the organism is stressed, some responses confer adaptive benefit; other responses are nonadaptive to environmental stresses. These morphoses (so named by Schmalhausen) are changes in the organism under stressful conditions, but they do not help the organism accommodate to that stress. For the example of Drosophila, Schmalhausen cites morphoses that include changes in the anatomy under conditions of increasing temperature, such as enlargement of the eye. The larger eye may give the fly no relief from heat, but may improve its getting around in dim light, an unrelated selective condition.

Taking together these two kinds of phenotypic changes, we would only know the organism's all-encompassing norm of reaction after it had been stressed in all combinations of conditions and durations of exposure. The span of responses would encompass the entire range of phenotypes the organism could generate from its single genotype; it would reveal the total latent phenotypic variation within the organism that could be generated without new genetic variation. This is certainly a broadening of Baldwin's view, including as it does not only the evoked developmental, physiological, and behavioral adaptations but the nonadaptive morphoses as well.

The adaptive reactions are more easily discussed, for they are used in the way Baldwin foresaw. When the organism enters or finds itself in a different environment, it adapts somatically to the extent it can. It survives, though stressed and perhaps marginally reproductive. Then heritable variants arise in the population that extend and stabilize the phenotypic adaptation, and they are selected. This is the Baldwin effect restated in more modern terms. After the stabilizing mutations, the organism is more reproductively fit and presumably less stressed. The change has been stabilized by internal heritable agencies rather than by external nonheritable ones, and the trait is produced each generation as part of the animal's embryonic development. A new norm of reaction has been generated, and bit by bit the organism can adapt to new circumstances.

The evolutionary significance of nonadaptive "morphoses" in the generation of phenotypic variation is harder to explain, except when the morphosis is fortuitously adaptive for some other selective condi tion. That is, one condition provokes or unmasks the morphosis, but a coincident, separate condition selects it.

In both the immediate adaptive and nonadaptive responses to environment, phenotypic change does not depend on new mutations or genetic variation. In the stabilization phase, the genetic variation may come from reassortment of existing variation in the population rather than from new mutations. The components and processes needed to produce the initial phenotypic variation are already there, without genotypic change. Before the organism meets the selective condition, its response is already encoded in its genome. Only "small" regulatory changes deriving from genetic variation in the population are needed to stabilize the change, bringing it under internal control. If the organism's "envelope of reaction" is vast, the organism as a whole can be described as a great exploratory system with many possible outcomes, from which random genetic change stabilizes a particular outcome relevant to the selective conditions. Thus, Schmalhausen simplified the thinking about phenotypic novelty by saying that it is within and we do not see it; it does not have to be created anew.

Experiments on the Baldwin Effect

At about this same time, Conrad Waddington in England began to think along similar lines and arrive at similar conclusions. His term for stabilizing selection was genetic assimilation, a term still in use, and he added two refinements as the result of experiments he performed with Drosophila. For example, he exposed fly populations to high salt in their food, to evoke their somatic adaptability toward salt; he then selected for increased salt tolerance. Or he exposed flies to ether during embryogenesis to provoke the later development of an extra pair of wings (four wings rather than two), a morphosis, after which he himself selected for flies with this outcome. (Extra wings seemed in no way to protect the fly against ether.) Or he shocked pupae at high temperature, which blocked the later development of cross-veins on the wings, after which he also selected for flies with this morphosis, as illustrated in

Figure 12. In all cases, he repeated the treatments and selections for 20-25 generations.5

First, he found, the initial population was usually quite heterogeneous in its response to the treatment, because of preexisting genetic variation and prior environmental conditioning. And second, much of stabilization, or genetic assimilation, was afforded by old genetic variation already in the population and brought into new combinations during successive matings, not by new mutations. The success of assimilation therefore depended on the genetic variation available in the population. With genetically heterogeneous populations, Waddington could, by the end of 20-25 generations of selection, reliably obtain populations showing the phenotypic novelty at high frequency, now independent of the special environmental conditions of treatment. With genetically homogeneous (inbred) populations, the effect did not appear.

Susan Lindquist is a geneticist who in the 1990s studied the resistance of organisms to heat. She extended the Waddington experiments to discern how heat unmasks cryptic phenotypic and genotypic variation. Excessive heat, like other stress conditions, causes most proteins of the cell to unfold and lose activity. The organism produces several kinds of special heat shock proteins (Hsp), or chaperone proteins as they are called, that guide the refolding of unfolded proteins back into their active form, thereby mediating recovery from heat. Hsp90 is one of these proteins.

As it turns out, even without heat, the Hsp90 protein is continuously important for folding newly made proteins correctly, especially large proteins of signaling pathways. When the organism is heated, the Hsp90 is recruited to refold damaged proteins, but there is not enough Hsp90 for folding the new proteins most in need of chaperone assistance. Aberrant phenotypes emerge—not just the cross-veinless kind pursued by Waddington but a wide range of others. Any of them could be selected by the researcher, and after some cycles of heating and selection would become stabilized. The spectrum of altered pheno-types differs in various stocks of flies, showing that genotypic diversity exists relative to which proteins are most dependent on chaperones.6

Figure 12 The Waddington experiment. A deliberate selection experiment demonstrating genetic assimilation after exposure to heat. By the twentieth generation, some flies developed cross-veinless wings even without heating. Note that the cross-veinless trait confers no benefit against heat.

Lindquist and her colleagues tested not just Drosophila but also Arabidopsis, a small flowering plant. They lowered Hspgo activity in several ways, by heat (as described above), by mutation of Hspgo itself, and by exposure of the organisms to a chemical agent that inhibits hspgo specifically. Many alterations of morphology were seen and, if desired, these could be put through stabilizing selection so that they would continue to form even after Hspgo was restored to full activity. The variety was great: different altered parts and different degrees of alteration. Lindquist refers to hspgo as a "capacitor for phenotypic variation." (Capacitors in electrical circuits accumulate a reservoir of charge and release the charge when there is a change in the circuit.)7

Several evolutionary biologists, including Mary Jane WestEberhard, along with Carl Schlichting and Massimo Pigliucci, emphasized the possible broad applicability of this adaptation-assimilation hypothesis in explaining the origin of complex phenotypic variations. In West-Eberhard's view, evolution of a novelty proceeds by four steps. In the first, called trait origin, an environmental change or a genetic change affects a preexisting responsive process, causing a change of phenotype (often a reorganization). At this initial step, she regards environmental stimuli as apt to be more important to evolution than genetic variation. The traits may or may not be adaptive; if they are not, they resemble Schmalhausen's morphoses or Waddington's and Lindquist's temperature-evoked changes.

In the second phase, the organism adapts or accommodates to its changed phenotype by compensating in part for the perturbed condition by using what we would say are its highly adaptive core processes.

In the third phase, recurrence, a subset of the population continues to express the trait, perhaps owing to the continued environmental stimulus.

In the final phase, genetic accommodation, selection drives gene frequency changes that increase fitness and heritability, although the phenotypic change is not necessarily ever completely under genetic control. While having a heritable component, it could retain an environmental dependence.

Thus this model, like Schmalhausen's, has a phenotypic accommodation phase followed by a genotypic accommodation phase. Most elements of a phenotypic novelty would not be new, and the role of mutations would be to provide small, heritable regulatory modulations rather than to create major innovations.8

Leading evolutionary biologists, including creators of the Modern Synthesis such as Ernst Mayr and George Simpson, were not impressed by adaptation-assimilation ideas. Some critics, missing the point, said that somatic adaptation is not heritable and hence is irrelevant to evolution, to which physiologists said the capacity for generating a broad range of somatic adaptations is as heritable as anything else. Other scientists acknowledged that the Baldwin effect "probably has occurred, but there is singularly little concrete ground for the view that it is a frequent and important element in adaptation."9

Simpson, the leading paleontologist of the Modern Synthesis, doubted that most traits could be regarded as stabilized somatic adaptations. He was most interested in anatomical changes, not physiological and behavioral alterations. Overheating animals might indeed create a few anomalies, such as larger eyes in fruit flies, but it was hard to imagine that it could elicit major anatomical innovations like jaws or wings. The somatic changes seemed quantitative, not qualitative— strengthening and weakening existing traits, not inventing new ones. In the middle of the twentieth century, evolutionary biologists tended to conclude that the Baldwin effect, if it exists, is of only minor relevance to evolution. Yet in a more modern molecular context, we believe that somatic adaptation, when applied to the conserved core processes, can help us resolve vulnerabilities in evolutionary theory about the origin of novelty.

A second reservation was that the adaptation-assimilation ideas, if correct, would have required the organism to hold much of future evolutionary adaptation within itself; it was hard to visualize how this broad potentiality would have been selected previously in evolution. The Baldwin effect seemed suitable for exploration of small deviations from the existing phenotype but not for radical new experiments. The organism might have a reaction norm for temperature, which would allow rapid evolutionary change within a certain range, but after exceeding that range, multiple characters would seemingly have to change at once. For physiological ranges that the organism might never see, such as the invention of new structures like eyes and wings, the organism in all likelihood would not contain within itself a plasticity extending to phenotypes it had never explored. Such reservations might not extend to morphoses, since these had not been previously selected for adaptability.

Was this article helpful?

0 0

Post a comment