Darwin and the Galápagos

Charles Darwin's historic visit to the Galápagos Islands in 1835 represents a landmark in the annals of science. But contrary to the legend long surrounding Darwin's famous Galápagos visit, he continued to believe that species were immutable for nearly a year and a half after leaving these islands. This delay in Darwin's evolutionary appreciation of the Galápagos evidence is largely owing to numerous misconceptions that he entertained about the islands, and their unique organic inhabitants, during the Beagle voyage. For example, Darwin mistakenly thought that the Galápagos tortoise–adult specimens of which he did not collect for scientific purposes–was not native to these islands. Hence he apparently interpreted reports of island-to-island differences among the tortoises as analogous to changes that are commonly undergone by species removed from their natural habitats. As for Darwin's finches, Darwin initially failed to recognize the closely related nature of the group, mistaking certain species for the forms that they appear, through adaptive radiation, to mimic. Moreover, what locality information he later published for his Galápagos finch specimens was derived almost entirely from the collections of three other Beagle shipmates, following his return to England. Even after he became an evolutionist, in March of 1837 (when he discussed his Galápagos birds with the eminent ornithologist John Gould), Darwin's theoretical understanding of evolution in the Galápagos continued to undergo significant developments for almost as many years as it took him to publish the Origin of Species (1859). The Darwin-Galápagos legend, with its romantic portrait of Darwin's 'eureka-like' insight into the Galápagos as a microcosmic 'laboratory of evolution', masks the complex nature of scientific discovery, and, thereby, the real nature of Darwin's genius.     

Reef fish behaviour during a total solar eclipse at Pinta Island, Galápagos

Diurnal fishes responded rapidly to decreased light levels during a total solar eclipse and sought shelter in the reef. Nocturnal fishes sometimes left the cover of their daytime habitats. The activity cycles of diurnal reef fishes appear to be a direct response to changes in light level.     

The microevolution of the Galápagos marine iguana Amblyrhynchus cristatus assessed by nuclear and mitochondrial genetic analyses

Marine iguanas may have inhabited the Galápagos archipelago and its former, now sunken islands for more than 10 million years (Myr). It is therefore surprising that morphological and immunological data indicate little evolutionary divergence within the genus. We utilized mitochondrial DNA (mtDNA) sequence analyses and nuclear DNA fingerprinting to re-evaluate the level and pattern of genetic differentiation among 22 marine iguana populations from throughout the archipelago. Both genetic marker systems detect a low level of within-genus divergence, but they show contrasting levels of geographical subdivision among the populations. The mitochondrial gene pools of populations from different regions of the archipelago are isolated, and the mtDNA pattern appears to follow the sequence in which the islands were colonized by marine iguanas. Conversely, the nuclear DNA study indicates substantial interpopulational gene exchange, and the geographical distribution of the nuclear markers seems to be determined by isolation by distance among the populations. The natural history of marine iguanas suggests that the contrasting nuclear and mitochondrial DNA patterns result from an asymmetric migration behaviour of the two sexes, with higher (active and passive) interisland dispersal for males than females. Separate genetic analyses for the sexes appear to support this hypophesis. Based on these findings, a scenario is proposed that explains the marine iguanas' low genetic divergence, notwithstanding their long evolutionary history in the Galápagos archipelago.     

The classical case of character release: Darwin's finches (Geospiza) on Isla Daphne Major, Galápagos

The history of how Darwin's medium ground finch (Geospiza fortis) on Isla Daphne Major, Galápagos became a textbook example of character release is reviewed. Four hypotheses for the evolution of the intermediate-sized Daphne fortis are examined, including genetic drift/founder effect, hybridization with the small ground finch (G'. fuliginosa) , food supply (or local adaptation hypothesis) and character release in allopatry.
Modern data suggest that genetic drift is unlikely to have been important, due to inadequate isolation and over-riding selection and introgression on Daphne. All three remaining hypotheses have probably played a role. Hybridization with G. fuliginosa occurs, although it cannot counteract the selection pressures seen during our study. Local adaptation has also occurred, with natural selection changing the relative frequencies of fortis phenotypes in response to changes in Daphne food supplies. The selection resulted from correlations between the size of seeds available, feeding behaviour and morphology. However, recent phenotypic tracking has resulted in larger, not smaller phenotypes. There is also evidence for character release in the form of diet expansion by G. fortis during periods of food shortage, and indirect evidence for interspecific competition between fortis and the cactus ground finch (G. scandens). The Daphne fortis phenotype probably represents a balance between introgression with fuliginosa, selection for larger body size in dry years and selection for smaller body size in wet years. The simple textbook account of a character shift caused by the accidental absence of competitors should be qualified to reflect the ecological complexity of the situation.     

A new look at evolution in the Galápagos: evidence from the late Cenozoic marine molluscan fauna

Endemism is not as common in the marine invertebrate fauna of the Galápagos Islands region as in the adjacent terrestrial biota. Marine invertebrates in the Galápagos are largely cosmopolitan species from the Panamic, Indo-Pacific, Californian, or Peruvian faunal provinces. However, an endemic component is also present in the fauna. The observed pattern among marine invertebrate organisms can be accounted for by at least two processes: (1) genetic continuity between mainland and island populations mediated through planktonic larvae; and (2) lower rates of intrinsic evolutionary change. The evolutionary scenario standardly applied to terrestrial organisms in the Galápagos, namely, adaptive radiation and speciation in reproductive isolation from mainland source populations, does not apply to all marine invertebrates. Evidence in support of the alternative scenario for marine invertebrates comes from both published records of species occurring in the islands and recent studies of fossil-bearing deposits on several islands in the archipelago. Two misconceptions–considering the islands and sedimentary deposits to be older than now thought, and equating the rate of evolution of the terrestrial biota with the marine biota–can lead to an incorrect interpretation of evolution in the Galápagos Contrasts between marine invertebrate and terrestrial organisms serve to illustrate some fundamental differences which have important evolutionary implications. Some of these are: endemism; dispersal; taxonomic relationships; island definitions; rates of evolutionary change; and age of fossils. In terms of Darwin's evolutionary scenario, terrestrial organisms represent the paradigm and marine organisms represent the paradox.     

The Role of Endangered Species Reintroduction in Ecosystem Restoration: Tortoise–Cactus Interactions on Española Island, Galápagos

We evaluated the role that endangered species reintroduction efforts can play in the larger context of ecosystem restoration. To do so, we examined interactions between endangered giant tortoises (Geochelone nigra hoodensis), currently being reintroduced to Isla Española, Galápagos, and an arboreal cactus (Opuntia megasperma var. megasperma), which is itself endangered and a keystone resource for many animals on the island. We collected information on spatial patterns of occurrence of cacti, tortoises, and woody vegetation and compared recruitment of juvenile cacti in areas occupied versus unoccupied by tortoises. Reintroduced tortoises appeared to suppress cactus recruitment near the few remaining adult cacti at the study site, but facilitate it at longer distances, with tortoise–cactus interactions mediated by the presence of woody vegetation, which likely alters tortoise movements and thereby patterns of cactus seed dispersal. The net effect of tortoises on cacti appeared to be positive insofar as tortoise presence was associated with greater recruitment of juveniles into cactus populations. Our study provides support for reintroducing endangered reptiles and other animals to aid ecosystem restoration in areas where they might once have played an important role in grazing upon and dispersing plants.     

Age-related changes in hematocrit in the Galápagos sea lion (Zalophus wollebaeki) and the Weddell seal (Leptonychotes weddellii)

In mammals, hematocrit (Hct) is optimized between the divergent requirements of blood flow characteristics and oxygen transport and storage capacity. This trade‐off plays a particularly major role in marine mammals, in which oxygen demand during sustained diving requires high Hct levels. Galápagos sea lions (Zalophus wollebaeki) need a long time after birth to develop from terrestrial life to the state of an independent forager at sea. We here show that pups were born with high Hct of 45%, then reduced Hct during the first 40 d of life to 31% while they remain constantly ashore, and increased Hct again until the adult level (57%) is reached at 1 yr of age when they begin to dive for foraging at sea. A similar, but much more rapidly changing pattern is seen in Weddell seal pups, but not in northern elephant seals, where no reduction in Hct is seen after birth. These and similarly changing patterns in terrestrial mammals likely reflect species specific functional adjustments during development due to a trade‐off between the costs of circulation and the changing need to store and transport oxygen.     

Flightlessness in the Galápagos cormorant (Compsohalieus [Nannopterum] harrisi): heterochrony, giantism and specialization

A study of flightlessness in the Galápagos cormorant (Compsohalieus [Nannopterum] harrisi) was undertaken using study skins and skeletons of C. harrisi and eight flighted confamilials; in addition, four skin specimens and disassociated skeletal elements of the extinct spectacled cormorant (C. perspicillatus) of Beringia, reputed by some to have been flightless, were studied. Anatomical specimens of C. penicillatus and C. harrisi were dissected for myological comparisons. Flightless C. harrisi is 1.6 to 2.2 times as heavy as its extant flighted congeners; males averaged 3958 g and females averaged 2715 g in total body weight. Estimates of body weight for C. perspicillatus based on femur length approximated 3900 g. Wing lengths of C. harrisi were smaller than those of any other cormorant, averaging 190 mm and 170 mm for males and females, respectively. Wing-loadings (g body mass.cm^-2 wing area) of flighted cormorants ranged from 1.0 to 1.7. Estimated wing-loadings, incorporating approximate wing areas, were 2.0 and 5.1 g.cm^-2 for C. perspicillatus and C. harrisi, respectively; the former suggests that C. perspicillatus was probably capable of laboured flight. The small wings of C. harrisi result from an c. 50% shortening of remiges, accompanied by reduced asymmetry of vane widths and increased rounding of the tips, and significant reductions in lengths of wing bones, particularly the radius and ulna. Numbers of primary and secondary remiges in C. harrisi remain unchanged. Multivariate morphometries revealed that sexual dimorphism in external and skeletal dimensions is significantly greater in C. harrisi than in flighted cormorants. Canonical analysis of six external measurements indicated that C. harrisi is distinguished primarily by its relatively short wings. Skeletal peculiarities of C. harrisi were diverse, including conformational changes in the sternum, furcula, coracoid, humerus, ulna, radius, carpometacarpus and patella. Mensural comparisons confirmed substantial reductions in elements of the pectoral girdle of C. harrisi, particularly the sternal carina, as well as the alar skeleton, especially the radius and ulna. Differential shortening of the wing elements resulted in significant differences in proportions within the wing skeleton. These unique skeletal proportions of C. harrisi, in addition to its great overall size, combine to produce an immense multivariate skeletal distance between C. harrisi and all confamilials. Sexual dimorphism in skeletal dimensions, in both total and size-corrected data, was 2–3 times greater in C. harrisi than in other phalacrocoracids sampled. Most pectoral muscles of C. harrisi were absolutely or relatively smaller than those of C. penicillatus, in spite of its larger body size. No muscles or parts thereof were lacking in the pectoral limb of C. harrisi, but a number of qualitative differences distinguished the musculature of the flightless species, including: an exceptionally tough skin involving a well-developed M. pectoralis pars abdominalis and M. latissimus dorsi interscapularis; a thin, medially obsolete and laterally extensive M. pectoralis pars thoracica; a weakly developed M. rhomboideus profundus consisting of a variably tendinous fascia invested with three fasciculi of muscle fibres; an extraordinarily thick, extensive M. obliquus externus abdominis, which, together with a unique cnemio-costal slip of smooth muscle, restricts the metapatagium through an anchoring of M. serratus superficialis metapatagialis; and the presence of a unique alular muscle named here as M. levator alulae. Fusions of the tendons of origin and insertion, respectively, of M. flexor digiti superficialis and M. flexor digiti profundus in C. harrisi, muscles derived from a common muscle primordium, and the retention of a carpometacarpal tendon of M. flexor carpi ulnaris cranialis constitute strong evidence of pectoral paedomorphosis in C. harrisi. Mensural comparisons quantified the reduction of pectoral muscles in C. harrisi and indicated that these reductions were especially pronounced in the distal musculature. Morphological characteristics of Phalacrocoracidae, together with the exploitation of localized marine food resources and weakly developed seasonal movements of Compsohalieus, may have predisposed the founding population of C. harrisi to flightlessness. Anatomical changes in C. harrisi are exceeded in degree among foot-propelled diving birds by those of only a few fossil flightless birds (e.g. Hesperomis, Chendytes). Many of the morphological peculiarities of C. harrisi are paedomorphic, although several are not attributable to developmental heterochrony. These morphological characters of flightless C. harrisi are considered with respect to locomotion, feeding ecology, reproduction and demography of the species, and are compared with those of other flightless carinates.     

Multiple source pools for Galápagos plant species richness: a critical analysis of the line of sight connectivity index

An index (C_i*E) combining the number of line‐of‐sight islands (C_i) within a radius i and target island elevation (E) has been proposed as an improved predictive model of plant species richness (S_t) in the Galápagos Archipelago. We examined this index critically and found that several major flaws preclude it from being a useful predictive tool for the archipelago. Although the number of collecting trips to an island was reported over 20 years ago to have substantial predictive value for reported plant species richness in the Galápagos Islands, this relationship was ignored in multiple regression analyses of the index. When we included the number of collecting trips in different multiple regression analyses of the index, C_i*E had less predictive power than collecting trips or ceased to be significant at all. Additionally, the strong significant relationship between elevation and area in the Galápagos Archipelago results in area having a major confounding influence on the C_i*E index. When elevation is removed from the C_i*E index, the predictive power of C_i is far less than area alone. Finally, the data used to construct and correlate the C_i*E index with (S_t) were based only on a subset of the islands and species lists that were incomplete or out of date. Species richness on islands can be related to the interaction of different factors, so development and testing of indices like C_i*E is not inappropriate. However, it is important to examine the interrelationships among the components of these indices thoroughly, and not ignore the effect of factors already known to have high predictive power. We propose several ways in which more meaningful indices of source pool(s) capacity can be constructed.     

Lineage identification of Galápagos tortoises in captivity worldwide

Ex situ conservation strategies may be substantially informed by genetic data, and yet only recently have such approaches been used to facilitate captive population management of endangered species. The Galápagos tortoise Geochelone nigra is an endangered species that has benefited greatly from the application of molecular and population genetic data, but remains vulnerable throughout its range. The geographic and evolutionary origins of 98 tortoises in private collections and zoos on three continents were identified using mitochondrial DNA (mtDNA) control region sequences and multi-locus microsatellite genotype data relative to a large database of representative samplings from all extant populations, including historical population allele frequency data for the Geochelone nigra abingdoni taxon on Pinta by way of museum specimens. All but six individuals had mtDNA haplotypes previously sampled, with the novel haplotypes identified as most closely related to robust populations on the islands of Santa Cruz and Isabela. Multi-locus genotypic assignments corroborated the results obtained from the mtDNA analyses, with 83.7% of individuals consistently assigned to the same locality by both datasets. Overall, the majority of captive unknowns sampled were assigned to the La Caseta Geochelone nigra porteri population, with no fewer than six individuals of hybrid origin detected. Although a purported Pinta individual was revealed to be of Pinzón ancestry, the two females currently housed with Lonesome George exhibited haplotypic and genotypic signatures that indicate that they are among the most appropriate matches for captive breeding. More generally, molecular approaches continue to represent important tools for assessing conservation value, minimizing hybridization and guiding management programs for preserving the distinctiveness of G. nigra taxa in captivity.     

Population size and trends of the Galápagos Penguin Spheniscus mendiculus

We applied a capture–mark–resight (CMR) method to estimate the population size of the Galápagos Penguin Spheniscus mendiculus . In 1999, we estimated 1198 individuals, with lower and upper 95% confidence limits of 1054 and 1403 individuals, respectively, and estimated that approximately 57% of the total population was counted in an annual census. Applying this estimate to the 2003 census, we estimated that the population size for the whole archipelago was 1351 individuals. We also applied the correction factor derived from the 1999 CMR data to other censuses carried out between 1970 and 2003 and estimated a maximum population of about 4000 individuals in 1971 when the highest numbers of Penguins were counted. Although the Penguin population size has fluctuated in the last 33 years, the overall trend shows larger populations in 1970–80 followed by relatively smaller populations and a slow recovery rate in 1983–2003. The data set also shows two major population declines (in 1983 and 1998), which are coincident with El Niño episodes.     

Freshwater nematodes of the Galápagos

Freshwater nematodes of the Galápagos archipelago were studied from four samples collected in 1988. Eighteen species, two of which are new to science, belonging to five orders, Enoplida, Dorylaimida, Mononchida, Monhysterida and Chromadorida, are described. Alaimus wittmeri n.sp. is characterised by having a smooth cuticle, a more posteriorly enlarged pharynx, a posteriorly situated amphidial fusus, obscure but anteriorly situated excretory pore, a mono-opisthodelphic ovary, and by the absence of sexual dimorphism in the lip region, the absence of pre-vulval uterine sac, the presence of males that have few (4–7) and weak ventromedian supplements and absence of mucro on tail tip. Monhystera floreanae n. sp. is distinguished by its females having a small body (L = 492–634 µm) with a narrow lip region, narrow posteriorly located amphidial openings, a small value for the c'-ratio (c' = 6.0–7.5), weak vaginal sphincter, by the absence of post-vaginal gland cells, absence of males and absence of crystalloid bodies. The difficulty in the use of crystalloid bodies as diagnostic character in the genus Monhystera is discussed. Two forms, a big form and a small form, which are similar except in measurements are reported for Achromadora micoletzkyi. SEM photos of ruptured cuticle of A. micoletzkyi and A. semiarmata show unique intra-cuticular structures. Comparison of these structures in the two species is made. Papilliform instead of setiform outer labial sensilla are also observed in A. micoletzkyi. The presence of the male and fine crystalloid bodies in Achromadora semiarmata, dorsal body pore in Prismatolaimus dolichurus and Prismatolaimus kenyensis, forty longitudinal cuticular ridges around the mid-body of Stomachoglossa pachyderma, crystalloid bodies in Ironus dentifurcatus are reported here for the first time. SEM photos of twelve species: Monhystera floreanae n. sp., Monhystera somereni, Alaimus wittmeri n. sp., Monhystrella hastata, Paramphidelus dolichurus, Eutobrilus annetteae, Ironus dentifurcatus, Chronogaster cameroonensis, Rhabdolaimus terrestris, Stomachoglossa pachyderma, Achromadora micoletzkyi and Achromadora semiarmata, are also presented.The richest nematode fauna was found in El Junco (San Cristobal), the only permanent freshwater lake on the archipelago. The nine species recorded from the lake are probably only a fraction of the total nematode fauna. The very recent man made reservoir La Toma has a relatively high diversity, due to its vicinity to El Junco. It is argued that the freshwater nematodes from San Cristobal probably arrived there through passive transport by birds.The freshwater bodies encountered on Floreana are very different from those on San Cristobal and so is their nematode fauna. Part of the species found on these islands may have been transported by land birds, others may have invaded the freshwater from terrestrial habitats.Apart from the two new species, the other ones have a wide to very wide distribution (cosmopolitan, southern hemisphere, ...), so that endemism of freshwater nematodes is very low, if it exists at all. This is due to easy dispersal capacities of nematodes on the one hand and the recent colonization (not more than 10 000, max. 40 000 years) on the other hand.     

Recent research on the evolution of land birds on the Galápagos

A decade of research on the evolution of Galápagos land birds is reviewed, and outstanding questions to be answered are highlighted. Evolutionary studies have been restric 1 almost entirely to the four species of mockingbirds and the 13 species of Darwin's finches. Long-term field studies have been initiated on representatives of both groups. Co-operative breeding has been discovered in the mockingbirds (and hawks).
Lack's (1945, 1947) monographic treatment of Darwin's finches has been largely upheld and extended by morphological, ecological, behavioural and biochemical studies. While the phylogenetic origins of Darwin's finches still remain uncertain, the major groupings of the finches have been confirmed by the results of protein polymorphism analysis. Fossils of Darwin's finches have been discovered recently: their potential for illuminating evolutionary change has not yet been realized. Three other major developments are (1) quantitative confirmation of the role of interspecific competition in the adaptive radiation, (2) experimental confirmation of the role of morphological and song cues in species recognition, and experimental evidence of their evolution in the speciation process, and (3) direct study of natural selection on heritable quantitative traits in a population, and identification of its causes. Continuing studies of population variation are likely to reveal the contemporary importance of selection, migration and hybridization, and thereby help us to more fully understand the causes of the adaptive radiation of Darwin's finches.     

Pollination biology of the Galápagos endemic, Tournefortia rufo-sericea (Boraginaceae)

Tournefortia rufo-sericea is an endemic member of the Galápagos angiosperm flora. Although not uncommon within the archipelago, its status is presently listed as vulnerable and, as such, a complete knowledge of its pollination biology may prove useful in preventing its decline. Pollination experiments, flower-visitor observations, nectar and fluorescence studies, as well as pollen : ovule ratio and pollen size studies were included in this investigation. The small, white flowers of this species set fruit via open pollination (81%), autonomous self-pollination (80%), diurnal pollination (80%), and nocturnal pollination (85%). Ants are the most common visitor to the flowers of this species, primarily during the day, whereas beetles and moths make visits at night. A small amount of nectar is presumably produced, as moths are often seen probing the corollas. Experiments with fluorescent dust did not support interflower pollen movement. The pollen : ovule ratio was 4972 : 1, which suggests xenogamy, and the mean pollen size was 21 µm. Despite the pollen : ovule ratio, it appears that this species exhibits a breeding strategy of facultative autogamy, in which the majority of flowers set fruit via autonomous self-pollination, but limited outcrossing may occur when pollinators are available. Conservation efforts for T. rufo-sericea should focus on the control or eradication of exotic species.  © 2007 The Linnean Society of London, Botanical Journal of the Linnean Society , 2007, 153 , 21–31.     

Lineage fusion in Galápagos giant tortoises

Although many classic radiations on islands are thought to be the result of repeated lineage splitting, the role of past fusion is rarely known because during these events, purebreds are rapidly replaced by a swarm of admixed individuals. Here, we capture lineage fusion in action in a Galápagos giant tortoise species, Chelonoidis becki, from Wolf Volcano (Isabela Island). The long generation time of Galápagos tortoises and dense sampling (841 individuals) of genetic and demographic data were integral in detecting and characterizing this phenomenon. In C. becki, we identified two genetically distinct, morphologically cryptic lineages. Historical reconstructions show that they colonized Wolf Volcano from Santiago Island in two temporally separated events, the first estimated to have occurred ~199 000 years ago. Following arrival of the second wave of colonists, both lineages coexisted for approximately ~53 000 years. Within that time, they began fusing back together, as microsatellite data reveal widespread introgressive hybridization. Interestingly, greater mate selectivity seems to be exhibited by purebred females of one of the lineages. Forward‐in‐time simulations predict rapid extinction of the early arriving lineage. This study provides a rare example of reticulate evolution in action and underscores the power of population genetics for understanding the past, present and future consequences of evolutionary phenomena associated with lineage fusion.     

THE DIET OF GALÁPAGOS SPERM WHALES PHYSETER MACROCEPHALUS AS INDICATED BY FECAL SAMPLE ANALYSIS

Fecal samples were collected while following sperm whales ( Physeter macrocephalus ) off the Galápagos Islands, Ecuador. They contained 133 upper beaks and 164 lower beaks of cephalopods. Analysis of the lower beaks suggests that the sperm whales fed primarily on three genera of cephalopods; Histioteutbis (62%), Ancistrocbeirus (16%), and Octopoteutbis (7%). The beak dimensions indicate that the cephalopods ranged in mantle length from 5 to 54 cm and in mass from 12 to 650 g. Fecal samples varied significantly between five study years and over different parts of the study area, but the number of beaks collected per sample did not correlate significantly with defecation rate (a measure of feeding success). Using beak material from fecal samples gives a biased estimate of sperm whale diet, reducing the frequencies of very small and very large cephalopods. However, all other available methods of assessing sperm whale diet also possess biases.     

Beak morphology and song features covary in a population of Darwin's finches (Geospiza fortis)

Animal mating signals evolve in part through indirect natural selection on anatomical traits that influence signal expression. In songbirds, for example, natural selection on beak form and function can influence the evolution of song features, because of the role of the beak in song production. In this study we characterize the relationship between beak morphology and song features within a bimodal population of Geospiza fortis on Santa Cruz Island, Galápagos. This is the only extant population of Darwin's finches that is known to possess a bimodal distribution in beak size. We test the hypothesis that birds with larger beaks are constrained to produce songs with lower frequencies and decreased vocal performance. We find that birds with longer, deeper, and wider beaks produce songs with significantly lower minimum frequencies, maximum frequencies and frequency bandwidths. Results from the analysis of the relationship between beak morphology and trill rate are mixed. Measures of beak morphology correlated positively with 'vocal deviation', a composite index of vocal performance. Overall these results support a resonance model of vocal tract function, and suggest that beak morphology, a primary target of ecological selection in Darwin's finches, affects the evolution of mating signals. We suggest that differences in song between the two modes of the distribution may influence mate recognition and perhaps facilitate assortative mating by beak size and population divergence.  © 2006 The Linnean Society of London, Biological Journal of the Linnean Society , 2006, 88 , 489–498.