Natural Selection & Speciation

Natural Selection — Deep Dive

Part 1 of 4 — Darwinian Fundamentals: Variation, Fitness & Adaptation

The Logic of Natural Selection

Natural selection is not a hypothesis about any particular trait — it is a deductive consequence of three conditions. If a population has:

1. Variation — individuals differ in a trait
2. Heritability — some of that variation is transmitted to offspring
3. Differential reproductive success — trait variants differ in how many surviving offspring they leave

then the trait distribution MUST shift across generations. Remove any one condition and selection produces no evolutionary change: a trait that varies but is not heritable (e.g., scars) cannot evolve, and a heritable trait with no fitness consequence changes only by drift.

Fitness = Reproductive Success

Fitness is measured in offspring, not in strength, longevity, or dominance. A short-lived organism that leaves eight fertile offspring is fitter than a vigorous one that leaves two. Survival matters only as a MEANS to reproduction — a component of fitness alongside mating success and fertility. "Survival of the fittest" is best read as "reproduction of the best reproducers."

Selection Acts on Phenotypes; Populations Evolve

Selection screens individual phenotypes — a predator catches the slow rabbit, whatever its genotype. But the evolutionary response is a change in allele frequencies in the population. Two consequences the MCAT tests:

• Individuals do not evolve. An organism's genotype is fixed at fertilization; it can acclimate physiologically, but acclimation is not evolution.
• Environmentally caused phenotypes (a bodybuilder's muscles, a sun tan) are invisible to evolution because they are not transmitted.

Adaptation

An adaptation is a heritable trait that increased in frequency BECAUSE it improved fitness in that environment.

An exaptation is a trait that evolved serving one function and was later co-opted for another: feathers likely spread for insulation and display before any bird flew; middle-ear bones derive from reptilian jaw bones. Exaptations matter because they answer "what good is half a wing?" — the intermediate stages were useful for something else.

Misconceptions to Kill on Sight

Where Variation Comes From

• Mutation — the only ultimate source of NEW alleles
• Recombination (crossing over in prophase I), independent assortment (metaphase I), and random fertilization — generate new COMBINATIONS of existing alleles each generation; this is why sex accelerates adaptation

Trade-Offs

Fitness components compete for the same energy budget: bright coloration attracts mates AND predators; large clutch sizes reduce per-offspring care; sickle-cell heterozygote protection comes packaged with disease risk in homozygotes. Passages often present a trait as "paradoxically" harmful — the resolution is almost always a trade-off in another fitness component or life stage.

Worked Example — Guppy Coloration as a Fitness Trade-Off

Passage-style problem. In Trinidadian streams, male guppies vary heritably in the brightness of their orange spots. Field surveys show: in pools ABOVE waterfalls (no large predators), males are brightly colored; in pools BELOW waterfalls (with predatory pike cichlids), males are drab. A researcher transplants 200 guppies from a high-predation pool to a previously guppy-free, predator-free pool. Within about 15 generations, average male brightness increases significantly.

Question 1 — Why does brightness increase after the transplant? Brightness is under opposing selection from two fitness components: females preferentially mate with brighter males (mating success), while predators preferentially eat them (survival). Removing predators removes the survival cost, so net selection becomes directional FOR brightness. The response across generations confirms the trait is heritable.

Question 2 — Did individual transplanted males become brighter? No. Individuals do not evolve; the DISTRIBUTION shifted because bright males left disproportionately many offspring each generation. Any single fish's genotype was fixed at fertilization.

Question 3 — Did the guppies brighten "because they no longer needed camouflage"? No — that phrasing implies foresight. Bright variants already existed (standing heritable variation); the new environment merely changed which variants reproduced most. If no bright alleles had been present, no amount of "need" would have produced them; the population would have waited on mutation.

Question 4 — What result would falsify the heritability requirement? If offspring brightness showed no resemblance to father brightness (e.g., brightness tracked diet alone), the generation-to-generation increase could not be evolutionary; brightness would be an acquired, environmentally induced phenotype, invisible to selection.

Darwinian Fundamentals 🎯

Key Takeaways — Part 1

• Selection follows necessarily from heritable variation plus differential reproductive success; remove either ingredient and the trait cannot evolve
• Fitness = relative reproductive success; survival and vigor are only means to that end
• Selection screens phenotypes; evolution is a change in population allele frequencies — individuals never evolve, and acquired traits are not inherited
• Adaptation = heritable trait that spread because it raised fitness in that environment
• No foresight, no need-driven change, no perfection: selection can only sort existing variation under trade-off constraints
• Mutation is the sole source of new alleles; recombination, independent assortment, and random fertilization multiply combinations

• Low-yield extras: exaptation = a trait co-opted for a new role after evolving for another (feathers for insulation and display before flight; middle-ear bones from reptilian jaw bones), which answers "what good is half a wing?"

Natural Selection — Deep Dive

Part 2 of 4 — Modes of Selection, Sexual Selection & Altruism

The Three Modes, With Their Classic Examples

Disruptive selection is the mode with speciation potential WITHIN a single population: if the two favored extremes also begin mating assortatively (like with like), the population can split — a thread picked up in Part 3.

Sexual Selection: Fitness Includes Mating Success

Sexual selection is differential reproductive success arising from competition for MATES. Two channels:

• Intersexual selection (mate choice): members of one sex (usually females, the higher-investing sex) choose among the other. Drives ornaments — peacock trains, bright plumage, courtship displays.
• Intrasexual selection (same-sex competition): typically male-male combat or display for access to mates. Drives weapons and size — antlers, horns, sexual dimorphism in body mass.

Why do costly ornaments persist when they reduce survival? Because fitness is reproduction, not survival. A train that costs a peacock some predation risk but multiplies his matings can raise net fitness. Choice for such ornaments can be maintained because ornaments serve as indicators of male condition or genetic quality — only a healthy male can afford the handicap. The result is a survival/mating trade-off, and sexual dimorphism is its visible signature.

A second route is runaway (Fisherian) selection: the female preference and the male ornament become genetically correlated, so each generation's choice amplifies both.

Altruism, Kin Selection & Hamilton's Rule

Behavior that lowers the actor's own reproduction while raising a recipient's (alarm calling, helping at the nest, sterile worker castes) looks paradoxical — until genes are counted through RELATIVES. Your relatives carry your alleles with predictable probabilities (the coefficient of relatedness r): offspring and full siblings $r = 0.5$; half-siblings, grandchildren, aunts/uncles-to-nieces/nephews $r = 0.25$; first cousins $r = 0.125$.

Inclusive fitness = own reproduction (direct) + effects on relatives' reproduction, weighted by r (indirect). Kin selection favors an altruistic act when Hamilton's rule holds:

$rB > C$

where B is the benefit to the recipient and C the cost to the actor, both in offspring-equivalents. Predictions that data confirm: alarm calls in Belding's ground squirrels are given overwhelmingly by females living near kin; help flows preferentially to closer relatives; sterile insect workers raise siblings rather than offspring.

The Group-Selection Caveat

"For the good of the species" is almost never the right explanation. A self-sacrificing allele that benefits the group is undermined from within by cheaters who accept benefits without paying costs — individual-level selection is usually the stronger force. Apparent group-beneficial behavior on the MCAT should be re-explained via kin selection ($rB > C$) or reciprocity, not species-level benefit.

Worked Example — Applying Hamilton's Rule to an Alarm Call

Problem. A ground squirrel spots a hawk. Giving an alarm call raises her own chance of being taken, at an expected cost of $C = 1$ offspring-equivalent. Staying silent lets the hawk take a nearby juvenile. Evaluate whether calling is favored by kin selection when the juvenile at risk is (a) her full sibling, whose survival is worth $B = 3$ offspring-equivalents, and (b) her first cousin, same benefit.

Setup. Hamilton's rule: calling is favored when $rB > C$.

(a) Full sibling, r = 0.5. $rB = 0.5 \times 3 = 1.5 > 1 = C$

Calling is favored: the actor loses 1 offspring-equivalent directly but recovers 1.5 through the sibling's survival — her inclusive fitness rises by 0.5.

(b) First cousin, r = 0.125. $rB = 0.125 \times 3 = 0.375 < 1 = C$

Calling is NOT favored; the indirect gain cannot cover the cost. For a cousin, the benefit would need to exceed $C/r = 1/0.125 = 8$ offspring-equivalents before calling paid.

Interpretation. This is the quantitative core of the famous quip about laying down one's life "for two brothers or eight cousins": with $B$ measured in lives-worth-of-offspring, $2 \times 0.5 = 1$ and $8 \times 0.125 = 1$ just balance the cost of one self.

The quip is the geneticist J.B.S. Haldane's.

Data connection. In Belding's ground squirrels, alarm calling is concentrated in females — the philopatric sex that lives surrounded by mothers, sisters, and daughters — while males, who disperse away from kin, rarely call. That sex difference is exactly what kin selection predicts and what a "predator-confusion benefits everyone equally" account does not.

Selection Modes & Social Evolution 🎯

Key Takeaways — Part 2

• Directional selection shifts the mean (finch beaks, peppered moths, antibiotic resistance); stabilizing trims variance (birth weight); disruptive inflates variance and can produce bimodality (seedcrackers)
• Disruptive selection + assortative mating is the population-level on-ramp to sympatric speciation
• Sexual selection: intersexual = mate choice (ornaments), intrasexual = same-sex competition (weapons, size); both produce sexual dimorphism
• Costly ornaments persist because fitness includes mating success; ornaments can be honest indicators of condition
• Kin selection: relatedness r discounts benefits to relatives; altruism favored when $rB > C$; inclusive fitness = direct + indirect components
• Relatedness ladder: offspring/full sib 0.5, half-sib/grandchild/niece 0.25, first cousin 0.125; "good of the species" explanations lose to cheater invasion — reframe via kin selection or reciprocity

• Low-yield extras: runaway (Fisherian) selection, in which a genetically correlated preference and ornament amplify each other, is a second route to elaborate ornaments; the "two brothers or eight cousins" quip is J.B.S. Haldane's

Natural Selection — Deep Dive

Part 3 of 4 — Speciation, Reproductive Isolation & Phylogeny

What Is a Species?

The biological species concept: a species is a group of populations whose members can interbreed and produce viable, FERTILE offspring, and which is reproductively isolated from other such groups. Its limits are themselves testable material: it cannot classify asexual organisms (bacteria), fossils (no mating test possible), and it strains where distinct species occasionally hybridize (many plants, some birds and canids).

Reproductive Isolating Barriers

Prezygotic barriers prevent mating or fertilization — no zygote ever forms:

Postzygotic barriers act after fertilization:

• Hybrid inviability — hybrid zygotes fail to develop or die young
• Hybrid sterility — hybrids live but cannot reproduce: the mule (horse × donkey) is robust but sterile, partly because mismatched chromosome sets cannot pair properly in meiosis

• The mule's count: horse 2n = 64, donkey 2n = 62, so the mule carries an odd 63.

• Hybrid breakdown — first-generation hybrids are fertile, but their offspring are feeble or sterile

Prezygotic barriers are "cheaper": no gametes are wasted on doomed offspring, so selection can strengthen them where hybrids fare poorly (reinforcement).

Geography of Speciation

Allopatric speciation: a physical barrier (river, mountain uplift, habitat fragmentation) splits a population; drift and divergent selection accumulate differences until interbreeding fails even on recontact. The most common and best-documented route.

Sympatric speciation: divergence WITHOUT geographic separation. Two credible mechanisms:

• Polyploidy — especially in plants: a chromosome-doubling error creates individuals (e.g., tetraploids) instantly isolated from diploid parents, because triploid offspring are sterile. Speciation in one generation.
• Disruptive selection + assortative mating — ecological niches favor opposite extremes and like mates with like (Part 2's seedcracker logic; cichlid fish and apple maggot flies are the usual cases).

Adaptive radiation: one lineage rapidly diversifies into many niche-specialized species — Darwin's finches, Hawaiian honeycreepers, and silverswords — typically after colonizing an island system or after competitors vanish.

Reading Evolutionary Trees

• Divergent evolution produces homologous structures: same ancestral origin, possibly different functions (human arm, whale flipper, bat wing — one forelimb skeleton).
• Convergent evolution produces analogous structures: similar function, independent origins (bird vs. insect wings; shark vs. dolphin body form). Similarity from convergence (or reversal) is homoplasy — it misleads tree-building if mistaken for homology.
• A cladogram's nodes are common ancestors; a monophyletic group (clade) is an ancestor plus ALL of its descendants; sister taxa share an immediate common ancestor.
• Branch order, not left-right position or "ladder height," carries the information. No living taxon is "more evolved" than another — all tips have had equal time since the root.

Tempo

Gradualism: change accumulates steadily. Punctuated equilibrium: fossil species show long stasis interrupted by geologically rapid change, often associated with speciation events. "Rapid" here means thousands of generations — still ordinary population genetics, compressed relative to stasis, not a different mechanism.

Punctuated equilibrium was proposed by Niles Eldredge and Stephen Jay Gould (1972).

Worked Example — Reading a Cladogram Without Getting Fooled

Passage-style problem. A cladogram of four taxa has this branching structure: the deepest node splits taxon D from the rest; the next node splits taxon C from the pair (A, B); the shallowest node splits A from B. Character data: all four taxa share trait 1; C, A, and B share trait 2; A and B alone share trait 3; B alone shows trait 4.

Question 1 — Which taxa are sister taxa? A and B: they share an immediate common ancestor (the shallowest node) that neither shares with C or D. C is the sister taxon of the CLADE (A + B), not of A alone.

Question 2 — Is the group (C, A, B) monophyletic? Is (D, C)? (C, A, B) is monophyletic — it contains their common ancestor at the second node and all of that ancestor's descendants. (D, C) is NOT: the most recent common ancestor of D and C is the root, and a group containing the root's descendants must include A and B too. Leaving descendants out makes a group paraphyletic or polyphyletic, not a clade.

Question 3 — Where did trait 3 arise? On the branch leading to the (A, B) ancestor — after the split from C, before A and B diverged. Trait 4 arose on B's own terminal branch.

Question 4 — A student concludes that "B is the most evolved taxon because it has the most derived traits and sits at the end of the tree." Both reasons fail. Every tip has experienced the same amount of time since the root, and each lineage has been evolving throughout — D's terminal branch is the LONGEST-running independent lineage, yet that does not make D "more evolved" either. Trait counts reflect which characters were chosen for the table, and left-right order on a cladogram is typographically arbitrary: rotating any node (drawing B before A) depicts the identical evolutionary history.

Question 5 — Suppose trait 2 were later found in taxon D but analysis shows it arose independently there. Then trait 2 in D is homoplasy (convergence): similarity NOT inherited from a common ancestor. Counting it as homology would wrongly pull D inside the (C, A, B) clade.

Speciation & Phylogeny 🎯

Key Takeaways — Part 3

• Biological species concept = interbreeding + reproductive isolation; breaks down for asexual taxa, fossils, and hybridizing species
• Prezygotic barriers (habitat, temporal, behavioral, mechanical, gametic) block zygote formation; postzygotic barriers (inviability, sterility — the mule — and breakdown) waste it
• Allopatric speciation: geographic split, then divergence; sympatric speciation: polyploidy (instant, plant-heavy) or disruptive selection with assortative mating
• Adaptive radiation: one colonist lineage → many niche specialists (Darwin's finches, honeycreepers)
• Divergence → homology (same origin, different function); convergence → analogy/homoplasy (same function, different origin)
• Cladograms: nodes = common ancestors, clade = ancestor + ALL descendants, sister taxa share the immediate node; branch rotation is free, and no tip is "more evolved"
• Punctuated equilibrium = long stasis + geologically rapid change at speciation; still standard mechanisms, different tempo

• Low-yield extras: the mule's 63 chromosomes (horse 64, donkey 62); punctuated equilibrium's proposers, Eldredge and Gould

Natural Selection — Deep Dive

Part 4 of 4 — MCAT Integration: Evidence, Experiments & Data Interpretation

The Lines of Evidence (and what each uniquely shows)

• Named transitional fossils: Tiktaalik (a fish with limb-like fins and a neck, intermediate between fish and tetrapods) and Ambulocetus (a legged, amphibious whale ancestor).

Three Datasets the MCAT Recycles

1. Antibiotic resistance and the Lederberg replica-plating logic. The key question: do resistance mutations arise IN RESPONSE to the antibiotic, or do they pre-exist? Replica plating answers it: colonies grown WITHOUT antibiotic are stamped onto antibiotic plates, and resistant colonies appear at the SAME positions on every replica. Since the antibiotic never touched the master plate, the resistance mutations must have arisen randomly BEFORE exposure — the drug selects pre-existing variants; it does not induce them. Any passage where a stressor "creates" adaptive mutations on demand is describing the wrong (Lamarckian) model.

2. The Grants' finches. On Daphne Major, the 1977 drought left mostly large, hard seeds. Ground finches with deeper beaks survived at higher rates; because beak depth is strongly heritable (offspring resemble parents), the NEXT generation's mean beak depth was measurably larger. All three ingredients were documented in the wild: variation, heritability, differential survival → directional evolutionary response within a few generations. Wet years with abundant small seeds later pushed the mean partway back — selection tracks the current environment, with no fixed direction.

3. Peppered moths. Dark morph frequency rose with industrial soot (birds ate conspicuous light moths on darkened trunks) and fell again after clean-air legislation — a reversible, environment-linked directional shift, replicated across regions.

Experimental Design Questions

To claim a trait is heritable and selected, an experiment needs:

• Heritability check: offspring-parent resemblance measured in a COMMON environment (common-garden or cross-fostering design), so family resemblance cannot be blamed on shared environment or diet.
• Selection check: fitness (survival AND reproduction) measured as a function of the trait, with a control population not exposed to the selective agent.
• Replication: multiple independent populations. Here is the selection-versus-drift discriminator in data (Part 2 of Population Genetics covers the drift side in depth): replicated, CONSISTENT directional shifts — in large and small populations alike — indicate selection; scattered shifts in both directions, worse in small populations, indicate drift.

Molecular Clocks (concept level)

Neutral mutations accumulate at a roughly constant average rate, so the number of neutral sequence differences between two lineages grows approximately linearly with time since their divergence. Calibrated against fossil-dated splits, sequence divergence estimates the TIMING of branch points. Caveats a passage may exploit: rates differ among genes and among lineages, and selection-constrained sites do not tick neutrally.

• Gene-rate extremes: histones (tightly constrained) tick very slowly; fibrinopeptides (little functional constraint) tick fast.

Worked Example — Interpreting a Resistance-Evolution Experiment

Passage-style problem. Researchers establish 6 replicate E. coli populations from a single antibiotic-sensitive colony. Populations 1-3 grow with a sublethal dose of streptomycin; populations 4-6 grow without antibiotic. Resistant cells per $10^6$ cells are counted at generation 0 and generation 200:

A replica-plating test on a SEPARATE, never-exposed culture finds resistant colonies at matching positions across replica antibiotic plates.

Question 1 — Do these data show the antibiotic caused resistance mutations? No. Populations 4-6 show that resistant mutants exist at low frequency (about $1$ to $2$ per $10^6$) even with no drug — mutation maintains them regardless of exposure. The replica-plating result clinches the timing: position-matched resistant colonies descend from mutants that arose on the master plate BEFORE any antibiotic contact. The drug's role in populations 1-3 was to change the FITNESS of pre-existing variants: $w_{\text{resistant}} \gg w_{\text{sensitive}}$ under treatment, driving the resistant fraction from about $10^{-6}$ to about $0.94$.

Question 2 — Why does the resistant fraction stay near 1 per million in populations 4-6? Without the drug, resistance typically carries a small fitness COST (resistant ribosomes work slightly worse), so selection weakly removes resistant cells while mutation keeps resupplying them — a mutation-selection balance, not an absence of mutation.

Question 3 — Which single result, if observed instead, would have supported induced (directed) mutation? Resistant colonies appearing at RANDOM, non-matching positions across replica plates only after antibiotic exposure, together with a resistant fraction in unexposed populations of exactly zero at all times. Matching positions are the fingerprint of pre-exposure mutational events propagated by cell division.

Question 4 — Is the rise in populations 1-3 selection or drift? All three independent replicates shifted massively in the SAME direction, in large populations — consistent, directional, size-independent change is the signature of selection, not drift.

Evidence & Experimental Reasoning 🎯

Key Takeaways — Part 4

• Evidence stack: fossils (transitional forms), comparative anatomy (homologous + vestigial structures), molecular homology, biogeography, and direct real-time observation
• Replica plating: resistance mutations pre-exist exposure (position-matched colonies from an unexposed master plate); antibiotics select, they do not instruct
• Grants' finches: documented variation + heritability + differential survival → measurable beak-depth evolution in the wild; direction reverses when the environment does
• Peppered moths: reversible, environment-linked directional selection replicated across regions
• Demonstrating adaptation experimentally requires a common-garden heritability check, fitness measured against the trait with controls, and replication
• Selection vs. drift in data: consistent, size-independent, replicated directional shifts = selection; scattered, size-dependent shifts = drift
• Molecular clocks: neutral divergence grows roughly linearly with time; calibrate with fossils, beware gene- and lineage-specific rate differences

• Low-yield extras: named transitional fossils (Tiktaalik, fish to tetrapod; Ambulocetus, legged whale ancestor); molecular-clock rate extremes (histones slow, fibrinopeptides fast)