* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Lecture 1
Survey
Document related concepts
Hologenome theory of evolution wikipedia , lookup
Organisms at high altitude wikipedia , lookup
Darwinian literary studies wikipedia , lookup
Theistic evolution wikipedia , lookup
Coevolution wikipedia , lookup
Inclusive fitness wikipedia , lookup
Evolutionary mismatch wikipedia , lookup
The eclipse of Darwinism wikipedia , lookup
Evolutionary landscape wikipedia , lookup
Punctuated equilibrium wikipedia , lookup
Saltation (biology) wikipedia , lookup
Transcript
Carlo Matessi IGM-CNR Pavia Italy 1 Laws of Adaptation Lectures on mathematical models of biological evolution by Carlo Matessi Lecture 1 Demography, or the “short-term dynamics” point of view Tuesday September 4, 15:30-16:30 University of Tennessee. Department of Ecology & Evolutionary Biology. Sep-Oct 2007 Carlo Matessi IGM-CNR Pavia Italy 2 Charles Darwin in the Galapagos islands The Galapagos are volcanic islands, formed around 5 million years ago, situated in the Pacific ocean about 1000 Km west of Ecuador M ap of the Galapagos Geographical position of the Galapagos downloaded from www.du.edu/~ttyler downloaded from www.geol.umd.edu When Darwin, in his trip around the world as a naturalist on board the survey ship “Beagle”, visited them in 1835 was amazed by their surprisingly diverse fauna which –besides a giant tortoise and three species of iguanas, two terrestrial and one marine– included a very diversified group of finches University of Tennessee. Department of Ecology & Evolutionary Biology. Sep-Oct 2007 Carlo Matessi IGM-CNR Pavia Italy 3 The Galapagos finches 1. Geospiza conirostris 2. Geospiza magnirostris 3. Geospiza fortis 4. Geospiza scandens 5. Geospiza difficilis 6. Geospiza fuliginosa 7. Cactospiza pallida 8. Platyspiza crassirostris 9. Camarhynchus pauper 10. Camarhynchus psittacula 11. Camarhynchus parvulus 12. Certhidia olivacea 13. Cactospiza heliobates The wide spectrum of bill sizes and shapes The large species complex of Galapagos finches downloaded from www.geolsoc.org.uk From BSCS, Biological Science: Molecules to Man, Houghton Mifflin Co., 1963 Several of these species inhabit distinct islands in the archipelago. None of these species exists in the mainland, where only one species is found that is distantly related to this complex. The diversity of bill sizes and shapes parallels an equally broad diversity of diet and habitat. University of Tennessee. Department of Ecology & Evolutionary Biology. Sep-Oct 2007 Carlo Matessi IGM-CNR Pavia Italy 4 Indications from Darwin’s finches Characteristic features of living beings are modified in (geological) time: biological evolution Most changes occur in such a way to enable idividuals to thrive in a given new environment and to exploit efficiently opportunities and resources specifically available in that environment: adaptation Adaptation Adaptation is manifested whenever structures or activities are distinctly adequate to achieve a specific end or to perform correctly a given duty. It is a distinctive property of living beings, where it is present under two aspects Internal adaptation: precise coordination and harmonious interaction between different parts of an organism at all levels of structure (molecular, subcellular, cellular, organs, systems of organs) External adaptation: tight correlation between characters of the organism and certain properties of the environment that are evidently important for its survival (e.g., food, shelter, etc.) University of Tennessee. Department of Ecology & Evolutionary Biology. Sep-Oct 2007 Carlo Matessi IGM-CNR Pavia Italy 5 Natural selection A process that can account for adaptive evolution of living beings It is necessarily active in any collection (population) of entities (individuals) where the following conditions are verified 1. entities have a limited life-span but self-replicate (by acquiring and transforming energy and extraneous materials from the outside) 2. self-replication is not exact, so that a limited amount of (transmissible) variation is present among replicates 3. length of life-span and rates of self-replication depend on (transmissible) features of the entities These three conditions are characteristic of (define) life, so that any population of organisms is permanently undergoing natural selection The question is whether this specific process can account for the features and properties of biological diversity To answer this type of questions we need tools to construct, based on the concept of natural selection, theoretical predictions about specific biological phenomena University of Tennessee. Department of Ecology & Evolutionary Biology. Sep-Oct 2007 Carlo Matessi IGM-CNR Pavia Italy 6 Short term evolutionary dynamics Time scale: generation time t ∈{0, 1, …} Object: a population of interbreeding individuals Properties: K genetically transmissible types Xα(t) = proportion of adult individuals of type α at time t Yα(t) = proportion of newborn individuals of type α at time t wα = probability that newborn of type α survive to reproduce fβγ(α) = mean number of offspring of type α born to pairs (β,γ) Assume: infinite population; random mating → newborn Y! (t + 1) = $$ X " # " (t)X # (t)f"# (!) $$$ X % " # adults " X! (t + 1) = (t)X # (t)f"# (%) w ! Y! (t + 1) # w" Y" (t + 1) " w(t) = mean fitness = $ $ $ w ! X" (t)X # (t)f"# (!) ! " # University of Tennessee. Department of Ecology & Evolutionary Biology. Sep-Oct 2007 Carlo Matessi IGM-CNR Pavia Italy 7 The case of one gene with n variants Gene variants (alleles): {A1, …, An} Types (genotypes): {A1A1, A1A2, …, AnAn} – first allele is received from mother, second from father ; K = n2 Xij(t) , Yij(t) = proportion of AiAj among adults , newborn of generation t pi(t) , qi(t) = frequency of allele Ai among adults , newborn p i (t) = ! j X ij + X ji 2 , q i (t) = ! j Yij + Yji 2 By the laws of genetics: fij | kl (ik) = fij | kl (il) = fij | kl ( jk) = fij | kl ( jl) = Fij | kl , fij | kl (!) = 0 " ! #{ik, il, jk, jl} Assume: Fij | kl = F/4 ∀ (ij | kl) Yij (t + 1) = ( 1 ! ! Xik (t)X jl (t) + Xik (t)Xlj (t) + Xki (t)X jl (t) + Xki (t)Xlj (t) 4 k l ( ) 1 ! ! ! ! Xik (t)X jl (t) + Xik (t)Xlj (t) + Xki (t)X jl (t) + Xki (t)Xlj (t) 4 i j k l ) " X (t) + X ki (t) % " X jk (t) + X kj (t) % Yij (t + 1) = $ ! ik '& $# ! '& # k 2 2 k ! Yij (t + 1) = p i (t)p j (t) " (ij) and q i (t + 1) = p i (t) " i University of Tennessee. Department of Ecology & Evolutionary Biology. Sep-Oct 2007 Carlo Matessi IGM-CNR Pavia Italy 8 Gene frequencies are sufficient state variables and satisfy the following recurrence equations: ! w p (t) ij p i (t + 1) = p i (t) j j !!w k kj p k (t)p j (t) = p i (t) w i (t) , " i = 1, … , n w(t) j w ij = w ji = probability that newborn A i A j survive to reproduce w i (t) = ! w ij p j (t) = mean fitness of allele i j w(t) = ! ! w ij p i (t)p j (t) = mean fitness i j Any equilibrium point of the above recurrence equations must satisfy: either p̂ i = 0 or w i = w ! i = 1, … , n University of Tennessee. Department of Ecology & Evolutionary Biology. Sep-Oct 2007 Carlo Matessi IGM-CNR Pavia Italy 9 Maximization of mean fitness (Fisher, 1930; Kingman 1961) Use a continuous time approximation by which the recurrence equations are transformed into differential equations Associate discrete generation indexes {0, 1, …, t, t+1, …} to equidistant points on a continuous time scale τ∈[0,∞): t ! " , t + k ! " + k# Moreover let: w ij = w + !vij Hence: w i (t) = w i (!) = w + "vi (!) , w(t) = w(!) = w + "v(!) , vi (!) = # vij p j (!) , v(!) = # # vij p i (!)p j (!) j j It follows that: p i (! + ") # p i (!) v (!) # v(!) = p i (!) i $ " w + "v(!) i dp i (!) v (!) # v(!) = p i (!) i as " $ 0 d! w Now we can take the time derivative of mean fitness: dp (!) dp (!) 2 dv(!) 2 2 = 2" " vij p j (!) i = 2" vi (!) i = " p i (!)vi (!) [ vi (!) # v(!)] = " p i (!) [ vi (!) # v(!)] $ 0 d! d! d! w i w i i j i and equality obtains only at the equilibria of the system The mean fitness is a Liapunov function for the gene frequencies dynamics: it increase in time and its stationary points with respect to the gene frequencies coincide with the equilibria of the system. Hence it reaches a maximum when the population attains a stable equilibrium University of Tennessee. Department of Ecology & Evolutionary Biology. Sep-Oct 2007 Carlo Matessi IGM-CNR Pavia Italy 10 Competition for limited resources The population dynamics of a community of species competing for common, limited resources can be described by a system of “Lotka-Volterra” differential equations S species - Nα(τ) = number of individuals of species α at time τ (continuous time) ' * dN ! (") = #! ) C! $ & B!% N% ("), N ! (") ; #! , C! > 0 , B!% - 0 . !,% dt ( + % !" C" = intrinsic rate of increase of species " C! = carrying capacity of species ! B!! B!" B!! = effect on species ! of competiton with species " Parameters Cα and Bαβ have been given in Theoretical Ecology a “microscopic” interpretation: resources vary in a 1-dimensional spectrum identified by a variable z∈(-∞,+∞) production function c(z) ≥ 0 ∀z: amount of resources of type z made available per unit time utilization function φα(z) ≥0 ∀z: harvesting, per unit time, unit resource and unit consumer, of resources type z, by consumers type α C! = $ c(z)"! (z)dz , B!# = B#! = $ "! (z)"# (z)dz University of Tennessee. Department of Ecology & Evolutionary Biology. Sep-Oct 2007 Carlo Matessi IGM-CNR Pavia Italy 11 Short-term evolution driven by competition The evolution of one, or more, species sharing limited resources, with respect to traits that affect their competitive ability can be represented in analogy with the “Lotka-Volterra” equations, by a slight extension of the “1-locus, discrete generations” evolutionary model considered sofar S species; trait variation due to one gene in each species, with ns alleles present in species s {A , … , A } = alleles in species s !{1, … ,S} s 1 s ns Structure of the community N sij (t) = number of newborn of species s and genotype Asi Asj at generation t Sufficient state variables N(t) = ! ! N sij (t) = total number of newborn in the community at generation t s u s (t) = p si (t) = ij !N s ij (t) = proportion of newborn of species s at generation t ij N(t) !" N % 1 j 2 s ij (t) + N sji (t) #$ %N s kl (t) = frequency of allele Asi among newborn of species s at generation t kl University of Tennessee. Department of Ecology & Evolutionary Biology. Sep-Oct 2007 Carlo Matessi IGM-CNR Pavia Italy 12 Demographic parameters 1 + Csij " # # # Bstijkl N klt (t) ! t k l ! = number of offspring born to each adult irrespective of species and genotype w sij = probability that newborn Asi Asj survive to reproduce = notice that Csij = Csji , Bstijkl = Bstjikl , Bstijkl = Bstijlk ts , Bstijkl = Bklij Recurrence equations N(t + 1) = N(t) + !N(t) #$C(t) " B(t)N(t) %& u s (t + 1) = u s (t) 1 ! 1 ! 1 ! p (t + 1) = p (t) 1 s i s i + Cs (t) " Bs (t)N(t) + C(t) " B(t)N(t) + Csi (t) " Bsi (t)N(t) ! + Cs (t) " Bs (t)N(t) , s = 1, … ,S , s = 1, … ,S , i = 1, … , n s where Csi (t) = ! p sj (t)Csij , Cs (t) = ! p si (t)Csi (t) , C(t) = ! u s (t)Cs (t) j i Bsi (t) = ! ! ! ! p sj (t)p kt (t)p lt (t)Bstijkl j t k l s , Bs (t) = ! p si (t)Bsi (t) , B(t) = ! u s (t)Bs (t) i s University of Tennessee. Department of Ecology & Evolutionary Biology. Sep-Oct 2007 Carlo Matessi IGM-CNR Pavia Italy 13 A quantity maximized by competition-driven evolution (Matessi and Jayakar, 1981) Switch from discrete time t to continuous time τ and approximate recurrence equations by differential equations t + k ! " + k# , $ = #r , let # ! 0 the differential equations dN(!) = rN(!) #$C(!) " B(!)N(!) %& d! du s (!) = ru s (!) #$Cs (!) " Bs (!)N(!) " C(!) + B(!)N(!) %& , s = 1, … ,S d! dp si (!) = rp si (!) #$Csi (!) " Bsi (!)N(!) " Cs (!) + Bs (!)N(!) %& , s = 1, … ,S , i = 1, … , n s d! Equilibria must satisfy C ! BN = 0 , assuming that the community can subsiste at all Cs ! Bs N = 0 , for all species that are not extinct Csi ! Bsi N = 0 , for all alleles that are not extinct, in non-extinct species University of Tennessee. Department of Ecology & Evolutionary Biology. Sep-Oct 2007 Carlo Matessi IGM-CNR Pavia Italy 14 A Liapunov function for this system: !(") = 2C(")N(") # B(")N(")2 In fact, as a function of time, Ψ is increasing because, as it can be shown, 2 2 2 d!(") = 2rN $%C # BN &' + 2rN ( u s $%Cs # Bs N # C + BN &' + 4rN ( ( u s p si $%Csi # Bsi N # Cs + Bs N &' ) 0 d" s s i and equality only obtains at the equilibrium points of the system Moreover, the stationary points –internal and on the boundary– of Ψ with respect to the state variables {N, us, pis} coincide with the equilibrium points of the system Hence Ψ is maximized in the course of competition-driven evolution University of Tennessee. Department of Ecology & Evolutionary Biology. Sep-Oct 2007 Carlo Matessi IGM-CNR Pavia Italy 15 The biological meaning of Ψ Use the “microscopic” interpretation of the interaction parameters: {z : !" # z # +"} : resource spectrum, the ensemble of resource types available in the environment c(z) ! 0 "z : production function, fixed for a given environment !sij (z) " 0 #z : utilization function of genotype Asi Asj in species s, the property that is subject to evolution Csij = " c(z)!sij (z)dz , Bstijkl = " !sij (z)!klt (z)dz !(z, ") = # # u s (")p si (")p sj (")!sij (z) : mean utilization function, changing in time s ij C(!) = ) c(z)"(!, z)dz , B(!) = ) "2 (!, z)dz , #(!) = ) %& 2c(z)"(!, z)N(!) $ "2 (!, z)N 2 (!) '(dz Define !(") = 2 c(z) # $(", z)N(") dz : square deviation from habitat productivity of the community harvesting rate [ ] % so that !(") = # c 2 (z)dz $ %(") Hence, the maximization of Ψ implies the minimization of Γ. Evolution maximizes the fit of the resource utilization pattern to the actual distribution of resouces in the habitat. University of Tennessee. Department of Ecology & Evolutionary Biology. Sep-Oct 2007