Darwin s Finches: A Thirty Year Study. I. Mit-DNA Based Phylogeny (Figure 1). 1. All Darwin s finches descended from South American grassquit (small finch) ancestor circa 3 Mya. 2. Galapagos colonized 1 st 3. Extant Cocos Island species derived from Galapagos species. Figure 1. Mt-DNA based phylogeny. P. inornata is the Cocos Island finch. II. Recall Allopatric Model of Speciation. 1. Division of originally panmictic population into isolates. 2. Morphological divergence in response to local selection. 3. Post-divergence sympatry & selection for reproductive isolation. Figure 2. Wallace model of Geospizid evolution.
III. Contrast with A. R. Wallace Peripatric Model (Figure 2). 1. ARW imagined successive invasion and ecological divergence produced existing species. 2. Implicitly assumed all islands and niches present ab initio. IV. Subsequent Realization: Finches and archipelago co-evolved Wallace-Grant model. Figure 3. Coevolution of Galapagos Islands and Geospizid finches. 1. Number of islands increased due to crustal spreading. 2. Climatic conditions changed with recurrent fluctuations (El Niño) superimposed on overall trend of increasing aridity. Figure 4. Changing bill shape (length/depth) ratio shape and body size. a. An initially warmer, moister climate may have favored long, narrow beaks for exploiting nectar and insects. b. Subsequent aridity may have favored alternative feeding strategies (seeds) and morphologies (deep beaks). 2
V. Bill Size Selection (Figure 5). 1. G & G confirmed famous British ornithologist David Lack s earlier (1947) conjecture that larger bills facilitate consumption of larger (harder) seeds. 2. G & G observed two episodes of selection on bill size both associated with droughts due to El Niño weakening. 3. 1977. Selection for large beaks in G. fortis. a. Population declined with small-beaked birds disappearing to a greater extent than large-beaked conspecifics. b. Not enough small seeds birds forced to eat medium and large seeds. 4. 1987. Selection for small beaks. a. This time plants producing large seeds were hit hardest. b. Small-beaked individuals favored over large. 5. In both drought years, there was less mortality in the cactus finch (G. scandens), which is less dependent on seeds. 3
Figure 5. (Left) G. fortis abundance (top) and seed abundance (bottom) from 1975 to 1978. Right. Selection for large beaks in 1977. Beak size (depth) distribution in 1976 and 1978. Birds with small beaks experienced greater mortality than their large-beaked conspecifics. 4
VI. Adaptive Landscapes (Figure 5). 1. Utilization and seed abundance used to construct maximum bird density as function of beak size. 2. No more than one species per peak. 3. Close association of optimal and observed beak sizes (Figure 6). VI. Hybridization. 1. Species within Geospizid genera can hybridize. 2. Hybrid individuals have intermediate beaks that may or may not reduce fitness as estimated in Figure 6. Figure 6. Estimated maximum densities vs. beak size for Geospiza spp. on 15 different islands. Circles, etc. indicate mean beak sizes of resident finch populations. 3. Principal barriers to hybridization in Geospizidae are visual and acoustical i.e., how the birds look and sound. 5
VII. Reproductive Isolation. 1. Mechanism is mate choice by. 2. If female preference based solely on appearance, ecological divergence would => reproductive isolation. 3. But song is also involved. Song evolution consequent to a. Passive changes in resulting from changing morphology (body size, syrinx volume, beak size). b. What females consider to be a good song 1. Figure 7. Air sacs and syrinx in birds. Air sacs function in respiration; the syrinx, in vocalization. VIII. River of Life Think River Delta. 1. Tree of life branches coalesce result of hybridization. 2. At its tips, the tree of life becomes a web. 1 Females appear to evaluate the songs of would-be mates vis-a-vis vocalizations that they heard when they were young, i.e., their father s. 6