Topic 17. Lecture 27. Evolution of Populations and Ecosystems-II

1. Topic 17. Lecture 27. Evolution of Populations and Ecosystems-II What questions can be addressed by considering Macroevolution of simple phenotypes? Independently evolving individuals: Gene transmission: 1. Phenotypic plasticity 1. Mutation 2. Non-interactive behavior 2. Maintenance of sex 3. Semelparity and iteroparity 3. Crossing-over 4. Clutch size 4. Systems of mating 5. Dormancy 5. Origin of sex 6. Aging 6. Outcomes of genetic conflicts Interactions between individuals:Complex population-level phenomena: 1. Warning coloration 1. Multicellularity and coloniality 2. Dispersal 2. Anisogamy and sex allocation 3. Aggression 3. Mate choice 4. Cooperation and altruism 4. Female preferences and male displays 5. Conflicts between gametes and sexes 6. Conflicts between relatives 7. Eusociality Today, we will consider the second half of these questions.

2. Interactions between individuals: 1) warning (aposematic) coloration Individuals that has good defenses often possess warning coloration that deters potential predators. When everybody within a population of dangerous or unpalatable prey has such coloration, this is an obviously beneficial phenotype, because predators quickly learn to avoid this prey. However, it is not obvious how can warning coloration originate by natural selection. Indeed, a single mutant with conspicuous coloration would be eaten soon, because the predators would not have a chance to learn that such coloration means trouble. In contrast, the origin of mimicry is not problematic: if members of some other species already have warning coloration that protects them, mimicking them must increase fitness. Examples of warning coloration: A wasp A salamander

3. A nudibranch gastropod A flatworm A skunk A frog

4. Data support the contention that warning signals are selected for their reliability as indicators of defense (American Naturalist 162, 377–389, 2003). The initial evolution of conspicuous warning signals presents an evolutionary problem because selection against rare conspicuous signals is presumed to be strong, and new signals are rare when they first arise. Several possible solutions have been offered to solve this apparent evolutionary paradox. Mathematical models indicate that selection on novel warning signals is number- rather than frequency-dependent. In most cases, there exists a threshold number of aposematic individuals below which aposematism is selected against and above which aposematism is selected for. The threshold number of aposematic individuals is lower when predation is intense. This threshold may be not high: if several sibs all have a warning coloration, this may be enough for this trait to spread (Evolution 60, 2246-2256, 2006).

5. Interactions between individuals: 2) dispersal Should an individual stay put or migrate? There may be situations when the fitness of an individual would be higher if it does not move (if the cost of migration is high), but an allele that causes an individual to move will nevertheless spread in the population. This paradox appears because if an individual does not migrate, it will be likely to compete, within the same local population, with related individuals. In contrast, an individual that moves to another population will face unrelated competitors, and its success would greatly increase the frequency of a migration-causing allele. Thus, migration can evolve even when it is never beneficial to an individual, both under asex and under sex (J. of Theor. Biol. 82, 205-230 1980). One can say that migrating to another place with some probability is an ESS (evolutionarily stable strategy). A simple ESS is a phenotype such that, if everyone in a population possesses it, a different phenotype cannot be advantageous and, thus, cannot invade. ESS is an important conceptual tool for analyzing complex frequency-dependent selection, such that fitness landscape depends in a complex way on the composition of the population.

6. Interactions between individuals: 3) aggression Aggression against members of the same population is very common. In different species, conflicts between individuals can have different forms and outcomes, including death of one or both opponents. Examples of aggressive behavior: Betta fighting fish Boxing" walnut flies

7. However, very often conflicts are "ritualized": neither opponent uses all the weapons available. This is particularly common in animals that possess deadly weapons. How could this "moderation" evolve? Can it be explained without invoking "bad for an individual but good of the species" group selection arguments? Examples of ritualized conflicts: Western diamondbacks Eastern orynxs

8. Simple game-theoretical considerations make it possible to address this and related questions. Consider only two behavioral phenotypes ("strategies") - hawk (H, always attack) and dove (D, always be nice). A pair of individuals contests benefit b. When two individuals fight, the winner gets benefit b, and the loser suffers the cost of injury c. Two H's will fight, and the expected payoff for each is (b-c)/2. Two D's will not fight and will split the benefit (they will roll a coin, perhaps), and the expected payoff for each is b/2. An H and a D will not fight (D runs away), so that H gets b and D gets 0. So, the payoff matrix is: My opponent: H D Me: H (b-c)/2 b D 0 b/2 So, what to expect, evolutionarily? If everybody is a D, H phenotype ("allele") is beneficial and will invade the population (of course). However, if everybody is am H, phenotype D is beneficial, as long as c > b (avoid fighting altogether, if it is too costly). Thus, evolutionarily stable strategy here is mixed - be a H sometimes and a D sometimes. If c >> b, the ESS is to play D most of the time. There may be better strategies that simple H and D or even their mixture: start fighting, but escalate a conflict only until some point. This probably explains the evolution of ritualized conflicts through individual (not group) selection. As always when the population-level approach is used, we ignore within-organism mechanisms that make individuals into hawks or doves.

9. Still a better strategy is to persecute a weak, and to run away from a strong - to take into account how dangerous your opponent is. OK, but do you want to honestly inform you opponent how dangerous you are? Guinea baboons appear to be honest in In contrast, a lizard Phrynocephalus mystaceus signaling their strengths. pretends to be more dangerous than it is. This is a complex subject, but, generally, honest signaling can evolve only if any signaling is costly - due to production cost or social cost of the signal. In general, signals of quality require high and differential costs to remain honest, that is, prevent low-quality cheaters from exploiting any fitness benefits associated with communicating high quality.

10. The highly variable black facial patterns of female paper wasps, Polistes dominulus serve as "badges of status". Facial patterns strongly predict body size and social dominance. In staged contests between pairs of unfamiliar wasps, subordinate wasps with experimentally altered facial features ('cheaters') received considerably more aggression from the dominant than did sham controls, indicating that facial patterns are signals and that dishonest signaling imposes social costs. If you try to pretend that you are stronger than you are, they will beat you up. Signals of social dominance attract more attention - so you better be strong if you display them (Nature 432, 218-222, 2004).

11. Interactions between individuals: 4) cooperation and altruism Individuals very often help each other, even when this is costly. Under what conditions can we expect costly cooperation to evolve? Consider a very simple model. There are only two phenotypes ("strategies") - cooperate (C) and defect (D). A plausible situation is for an individual the best outcome of a pairwise encounter occurs when it defects and its partner cooperates, and the worst outcome occurs when both partners defect. For example, if one partner (prisoner) defects (talks to with cops) and another cooperates (remains silent), D is released and C gets 10 years. If both defect (talk), each gets 7 years, and if both cooperate (remain silent), they are released after one year. So, the payoff matrix is: My opponent: C D Me: C -1 -10 D 0 -7 So what is better - to cooperate with your partner or to defect? This is a famous Prisoner's Dilemma. If the partners encounter each other only once, it is always better to defect - no matter what your partner does, your payoff would be higher if you do so.

12. However, if the same two partners encounter each other many times (repeated Prisoner's Dilemma), this conclusion is no longer valid. Indeed, if constant cooperation of both partners can somehow be established, both partners will benefit, as compared to the case of both of them constantly defecting: there will be only 1 year in jail, instead of 7 years, for each crime. Now, our strategies (phenotypes) become algorithms: on the basis of what it and its partner did in previous encounters, and individual has to decide whether to cooperate or defect at this time. It is impossible to investigate the whole space of possible algorithms. However, it has been established experimentally that a very simple phenotype that does very well in repeated Prisoner's Dilemma: start from cooperation, and afterwards do what you partner did in the previous round (tit-for-tat). TFT individual: CCCCCCDDCCCCCCDDDDCCC Its partner: CCCCCDDCCCCCCDDDDCCCC Perhaps, it may be even better to be more generous and to forgive occasional defections, so that a partner can correct a random error without breakdown of cooperation. Thus, costly cooperation can evolve by natural selection if a phenotype that practices it has the highest fitness. There are five situations when this is possible (Science 314, 1560-1563, 2006): 1) kin selection 2) direct reciprocity 3) indirect reciprocity 4) network reciprocity 5) group selection.

13. A cooperator is someone who pays a cost, c, for another individual to receive a benefit, b. A defector has no cost and does not deal out benefits. Cooperation cannot evolve in any mixed population, because defectors have a higher average fitness than cooperators. After some time, cooperators vanish from the population. However, a population of only cooperators has the highest average fitness, whereas a population of only defectors has the lowest. Thus, natural selection constantly reduces the average fitness of the population. Fisher's fundamental theorem does not apply here because selection is frequency-dependent: the fitness of individuals depends on the frequency of cooperators in the population. We see that natural selection in well-mixed populations needs help for establishing cooperation.

14. Five mechanisms for the evolution of cooperation: 1) Kin selection operates when the donor and the recipient of an altruistic act are genetic relatives. 2) Direct reciprocity requires repeated encounters between the same two individuals. 3) Indirect reciprocity is based on reputation; a helpful individual is more likely to receive help. 4) Network reciprocity means that clusters of cooperators outcompete defectors. 5) Group selection is the idea that competition is not only between individuals but also between groups.

15. Kin Selection "I will jump into the river to save two brothers or eight cousins" (Haldane). More precisely, kin selection can lead to the evolution of cooperation if the coefficient of relatedness between the interacting individuals, r, exceed the cost-to-benefit ratio of the altruistic act: r > c/b (Hamiltons' rule). Here, relatedness is defined as the probability of sharing an identical-by-descent allele. The probability that two brothers share the same gene by descent is 1/2; the same probability for cousins is 1/8. Direct Reciprocity Repeated Prisoner's Dilemma is an example of this situation. Indirect Reciprocity Helping someone establishes a good reputation. Interacting with somebody who has a good reputation is beneficial, thus, such individuals can be "rewarded". Network Reciprocity A cooperator pays a cost for each neighbor to receive a benefit. Defectors have no costs, and their neighbors receive no benefits. In this setting, cooperators can prevail by forming network clusters, where they help each other. Group Selection If small groups of individuals are units of selection, cooperation can evolve, because groups of cooperators have higher fitness. Some other possibilities There are also some possibilities. For example, cooperators can recognize each others ("Green beard model"). Perhaps, several of these mechanisms contribute to evolution of cooperation in nature (J. Evol. Biol. 20, 415-432, 2007).

16. Not only complex animals can cooperate. The social amoeba Dictyostelium discoideum is a model for social evolution and development. When starving, thousands of the normally solitary amoebae aggregate to form a differentiated multicellular organism known as a slug. The slug migrates toward the soil surface where it metamorphoses into a fruiting body of hardy spores held up by a dead stalk comprising about one-fifth of the cells (Behav. Ecol. 18, 433-437, 2007).

17. Complex population-level phenomena: 1) multicellularity and coloniality Multicellularity evolved many times independently, in red algae, brown algae, green algae, fungi and animals (not counting several other "borderline" cases). In green algae, it evolved several times independently. Volvox sp. Ulva sp. Chara sp. The origin of multicellularity and coloniality is a complex and murky subject. It seems that the correct way of thinking about this subject is to consider conflicts and cooperation among individual cells (or organisms) (J. Evol. Biol. 19, 1406–1409, 2006). During origins of multicellularity, kinship and genetic and behavioral structure depend upon the mode of group formation (for example, do cells aggregate or do they remain together after repeated cell divisions), and, if groups are formed from a propagule, on the size of that propagule.

18. Multicellularity opens a possibility of conflicts between selection at different levels. A dominant mutation causing Apert syndrome is much more common in children of older (>45 years) fathers. Apparently, this is because cells that carry this mutation have a selective advantage within male germline (PNAS 105, 10143-10148, 2008). The same is probably the case for achondroplasia. Hi, my name is Frans Wallenberg and I have Apert Syndrome. When I was child, maybe 3-4 years old, I got my first surgery for my fingers... www.apert.org/wallenberg/index.html

19. Complex population-level phenomena: 2) anisogamy and sex allocation Anisogamy (including its extreme form, oogamy) evolved many times. Isogamy Anisomagy Oogamy Why are sperm small and eggs large? The most plausible explanation is disruptive selection on gamete size: small gametes are favored because many can be produced, whereas large gametes contribute to a large zygote with consequently increased survival chances. This model assumes that increases in zygote size confer disproportional increases in fitness. It therefore predicts that increases in adult size should be accompanied by stronger selection for anisogamy. Data from the green algal order Volvocales upheld predictions that larger organisms should (i) have a greater degree of gamete dimorphism and (ii) have larger eggs (Proc. Roy. Soc. Lond. B 268: 879-884, 2001).

20. When two sexes (exogamous classes of gametes) are present resources are usually allocated to them, by all individuals in the population, in equal proportion. This 1:1 sex allocation evolves owing to the simple fact that a zygote gets 50% of genes from the mother, and another 50% from the father. As a result, 1:1 sex allocation is an EES in simple cases (Fisher 1930). It is easy to understand why this is the case. Suppose that sex ratio in the population is female-biased. Consider a rare genotype that produces more males than others. Because a son will transmit more genes than a daughter (there is a deficit of males, so each male transmits on average more genes than a female), in the 2nd generation this genotype will be overrepresented ("grandchild argument"). If there are 4 females for each male, a male, on average, transmits 4 times more genes than a female. More complex situations are possible in structured populations, or when reproductive success depends differently on the body size for males and for females. For example, if matings mostly occurs within a sibship, an EES is to produce just one male who will mate will all his sisters. Melittobia digitata, a small parasitic wasp with female:male ratio 20:1. A host larva is usually parasitized by just one female, so that only full sibs can mate each other.

21. Complex population-level phenomena: 3) mate choice Why should a female care with whom to mate? Here "female" means the sex that invests more into one mating. Mate choice can exist either due to an immediate benefit or to a delayed benefit of producing offspring with better genotypes. There are 2 feasible immediate benefits of mate choice: 1) Direct investment - it is better to mate with a vigorous male, who will help to rise kids. 2) Reducing harm - it is better to mate with a male who will do you less harm.

22. However, immediate benefits can hardly explain all the instances of mate choice. Indeed, a father often contributes nothing but genes to his offspring and a variety of mates can be harmless. Thus, delayed benefits, due to production of better offspring, are probably important. Such benefits can be of two, not mutually exclusive, kinds: 1) Sexual selection - a choosy females produces more attractive sons, because their father was more attractive ("Fisherian runaway"). 2) Nonsexual selection - a choosy female produces offspring with generally superior genotypes, because their father had good genes.

23. However, both these ideas are not without problems (Ann. Rev. Ecol. Syst. EVOLUTION, 37: 43-66, 2006): 1) (Fisherian runaway) mechanism is very fragile and does not work if there is even a slight cost of choice for a female. 2) Fisher's Fundamental Theorem implies that, at equilibrium, there should be no heritable variation in fitness and no correlation between the quality of a father and his offspring. However, data demonstrate that heritable variation and parent-offspring correlations in fitness within natural populations are often quite large, probably, due to never-ending influx of deleterious mutations. Thus, it seems that the ability of good-quality fathers to sire good-quality children is the main reason for the evolution of female mate choice, although this issue is not yet settled.

24. Complex population-level phenomena: 4) female preferences and male displays Quite often, females not only choose mates, but do it according to rather bizarre criteria. Indeed, why should anybody want to mate with a peacock? Clearly, such exaggerated and costly sexual displays can evolve only as a result of coevolution with female choice.

25. Costly female preferences for males with exaggerated traits that reduce viability can evolve when the exaggerated trait, although maladaptive per se, indicates high overall quality of the male's genotype. The following evolutionary scenario appears to be plausible: 1) Initially, high-quality males have slightly longer tails. 2) Females start using long (but still optimal) tails as a clue for choosing high-quality mates. 3) This females choice causes all males to evolve exaggerated tails, that reduce their fitnesses. However, high-quality males can tolerate longer tails. As a result, a stable female preference for very long tails, and stable exaggeration of tail length over viability optimum can evolve (Proc. Roy. Soc. Lond. B 269, 97-104, 2002). This scenario is supported by some data, but the issue is not yet settled.

26. Complex population-level phenomena: 5) conflicts between gametes and sexes An egg and a sperm have rather different "interests" - in the sense that different things must happen for an egg and a sperm to contribute its genes to the next generation. An egg needs to be fertilized - but only once, as otherwise it will not develop properly. A sperm needs to fertilize an egg, and has nothing to lose if it fertilizes an egg together with other sperms. Consequently, an evolutionary arms race can occur between the ability of a sperm to penetrate an egg and an ability of an egg to make sure that only one sperm will do this. After one sperm gets in, the whole egg envelop must instantly become resistant to all other sperms. Such conflicts are common and often lead to extremely rapid coevolution, within the same genome, of genes with egg- and sperm-specific expression. Protein lysin in red abalone (Haliotisrufescens) is responsible for sperm-egg interaction and evolves extremely rapidly.

27. Complex population-level phenomena: 6) Conflicts between relatives Evolutionary interests of genes expressed in relatives can often be very different (only if reproduction is sexual, of course, as otherwise all relatives have the same genotype). Such conflicts can lead to complex phenomena if neither side is in complete control. Both parent-offspring and sib-sib conflicts are common. The extreme form of sib-sib conflict is siblicide. If the amount of resources is insufficient to support all sibs, killing others may be the only chance for an offspring to survive. In this situation, siblicide may be also in the evolutionary interest of parents (i. e., may increase the efficiency of transmission of their genes), as otherwise they would have no surviving offspring. Of course, under other conditions parents may be interested in raising all their offspring (if there are enough resources), and older offspring can help incubating their younger sibs (this trait can evolve due to kin selection). Siblicide in the brown booby

28. There may also be conflicts between parents and offspring, if a parent continues to invest resources into offspring after fertilization. Indeed, the evolutionary interest of a parent is not to waste resources on weak offspring. In contrast, the evolutionary interest of each offspring may be to survive. In organisms in which developing embryos are independent of the mother, under optimal conditions over 99% of embryos can develop successfully. In contrast, in mammals and seed plants the success rate is much lower, which is counterintuitive, because an embryo is protected and supplied by the mother. In humans, at least 30% of pregnancies are spontaneously terminated at very early stages. >99% success rate ~70% success rate A possible reason is that the maternal organism refuses to support embryos that appear to be weak or abnormal. Perhaps, this effect diminishes with the maternal age, because of diminishing chances of having other children. This may be partially responsible with the increased frequency of birth defects with maternal age.

29. Complex population-level phenomena: 7) eusociality Eusociality is an extreme form of altruism, such that many individuals do not reproduce and, instead, help their relatives to raise their offspring. Often, only one female (queen) reproduces in a colony, with other individuals being sterile workers. Eusociality independently evolved 11 times in Hymenoptera, and originated sporadically in other animals. Queen and workers of honey bee, Queen and workers of an ant, Apis mellifera Formica fusca In hymenopterans, workers in a single-queen colony are her sisters and daughters.

30. Queen and workers in a termite. All modern species of termites (order Isoptera) are eusocial. Eusocial sponge-dwelling snapping shrimp, Synalpheus regalis. They live in colonies with tens to hundreds of members and only one reproductive female. Naked mole rat (Heterocephalus glaber) is a eusocial mammal. The queen is the only reproductive female in the colony.  Other individuals serve particular societal roles, such as soldiers and cleaners. Evolution of eusociality apparently proceeds through kin selection. In hymenopterans, who have haploid males, it may be aided by closer relatedness of a female to her sisters (75% of identical by descent genes) than to her daughters (only 50% of such genes).

31. The key factor in the evolution of eusociality appears to be monogamy. This is to be expected in eusociality evolved due to kin selection, because monogamy ensures high relatedness of daughter of the same mother. Phylogeny of eusocial Hymenoptera (ants, bees, and wasps). Each independent origin of eusociality is indicated by alternately colored clades. Clades with high polyandry (>2 effective mates) are in solid red, those with low polyandry (>1 but <2 effective mates) are in dotted red, and monandrous genera are in black (Science 320, 1213 - 1216, 2008).

32. Evolution of Ecosystems A ridiculously short summary Ecosystems consist of populations of many species. Obviously, properties of an ecosystem are affected by the evolution of constituent populations. There is no selection for the "well-being" of an ecosystem - instead, evolution of an ecosystem is a by-product of Darwinian evolution within the constituent populations. Such evolution can easily produce unexpected results even in very simple ecosystems. For example an ecosystem consisting of just two species, a prey and a predator, can exist either in an equilibrium state (if the predator is inefficient) or in a sate of stable oscillations (if the predator is more efficient). If the predator is very efficient, the amplitude of these oscillations can become very wide, which can lead to extinction of the predator, due to lower critical density phenomenon (Allee effect). Thus, adaptive evolution of a predator can first destabilize the ecosystem and later even lead to the predator's extinction (Darwinian extinction, Am. Nat. 61, 181-191, 2003; Nat. Rev. Genet. 8, 185-195, 2007). Possible consequences of slow evolution of more efficient predators.

33. Quiz: What observations and experiments can be used to establish the mechanisms of evolution of cooperation?