Origin and cross-century dynamics of an avian hybrid zone

Abstract

Background: Characterizations of the dynamics of hybrid zones in space and time can give insights about traits and processes important in population divergence and speciation. We characterized a hybrid zone between tanagers in the genus Ramphocelus (Aves, Thraupidae) located in southwestern Colombia. We evaluated whether this hybrid zone originated as a result of secondary contact or of primary differentiation, and described its dynamics across time using spatial analyses of molecular, morphological, and coloration data in combination with paleodistribution modeling.

Results: Models of potential historical distributions based on climatic data and genetic signatures of demographic expansion suggested that the hybrid zone likely originated following secondary contact between populations that expanded their ranges out of isolated areas in the Quaternary. Concordant patterns of variation in phenotypic characters across the hybrid zone and its narrow extent are suggestive of a tension zone, maintained by a balance between dispersal and selection against hybrids. Estimates of phenotypic cline parameters obtained using specimens collected over nearly a century revealed that, in recent decades, the zone appears to have moved to the east and to higher elevations, and may have become narrower. Genetic variation was not clearly structured along the hybrid zone, but comparisons between historical and contemporary specimens suggested that temporal changes in its genetic makeup may also have occurred.

Conclusions: Our data suggest that the hybrid zone likey resulted from secondary contact between populations. The observed changes in the hybrid zone may be a result of sexual selection, asymmetric gene flow, or environmental change.

Keywords: Andes; Cline; Distribution modeling; Hill function; Hybridization; Moving hybrid zone.

Fig. 1

Phenotypic variation in male specimens collected along the Ramphocelus flammigerus hybrid zone in southwestern Colombia. Individuals 1-6 correspond to R. flammigerus icteronotus (yellow-rumped form) from the plains of the Pacific coast (sector 1; see Fig. 2). On the other extreme, individuals 11-14 correspond to R. flammigerus flammigerus (scarlet-rumped form) distributed towards the Cauca River Valley (sector 3). Individuals 7-10 are intermediates collected near the center of the hybrid zone (sector 2)

Fig. 2

Study transect encompassing the Ramphocelus hybrid zone from the Pacific lowlands to the eastern slope of the Cordillera Occidental of the Colombian Andes. Blue dots indicate collection sites of historical specimens and purple dots indicate collection sites of current specimens sampled for phenotypic/genetic variation. The extent of each of the three sectors we defined in the hybrid zone (S1, S2 and S3; see text) is indicated by black dots on the transect

Fig. 3

Potential distributions of R. flammigerus predicted by MaxEnt using climatic data. Dark gray areas show suitable climatic conditions for the occurrence of flammigerus and icteronotus (a) 6000 years ago and (b) 21,000 years ago (LGM) as estimated employing niche models including all locality data. Light gray depicts climatically suitable areas for their occurrence at present. Note the smaller predicted range during the LGM and that the two forms likely did not exhibit a continuous range along the transect (dotted line) at that time, relative to the more extensive and continuous range modeled for 6000 y.a. and under current conditions. Potential distributions are also shown separately by taxon, as predicted by models excluding intermediate specimens and considering only localities of icteronotus (c and d) and flammigerus (e and f). Note that for both taxa, distributions likely spanned the hybrid-zone transect 6000 years ago but not 21,000 years ago; in the LGM, the potential distribution of flammigerus (f) was dramatically reduced

Fig. 4

Genealogical relationships of specimens of R. flammigerus showing limited geographic structuring and relatively low levels of sequence divergence among haplotypes. The phylogenetic tree on the left depicts relationships among nearly complete sequences of the cytochrome b gene obtained for individuals from the hybrid zone and other localitites inferred using maximum-likelihood (outgroups not shown); bootstrap values on nodes are shown when ≥ 50%. Colored circles indicate a qualitative assessment of the rump color (yellow, orange and red as in individuals 1-6, 7-10 and 11-14 in Fig. 1, respectively) and location in the hybrid zone (cyan, sector 1; green, sector 2; dark blue, sector 3) of individuals from the study transect exhibiting each haplotype. The numbers correspond to specimen identifications in Additional file 1: Table S1; all numbers refer to specimens from the hybrid-zone transect unless otherwise noted. Localities outside the transect in different provinces of Colombia (CO), Ecuador, and Panama are indicated with squares. Haplotype networks on the right focused on specimens from the hybrid zone show that genetic variation in 210 bp of the cytochrome b gene was not clearly consistent with position of individuals along the hybrid zone in 1956 (top) or in the present (bottom), although analyses of molecular variance suggest differences in patterns of genetic structure across time periods (see text); circle sizes are proportional to the number of individuals sharing each haplotype

Fig. 5

Estimates of population sizes over time obtained using the extended Bayesian skyline plot method applied to cytochrome b sequence data suggest demographic expansion towards the present in R. flammigerus. In each plot, median and credibility interval values are shown in black solid line and dashed lines, respectively. Blue lines correspond to 1000 genealogies used to estimate the 95% highest posterior density of population sizes. Bars in the histograms are proportional to the number of genealogies with values in the specific time interval. Analyses are shown for (a) all specimens (including flammigerus, icteronotus and intermediate individuals) and separately for specimens assignable to (b) icteronotus and (c) flammigerus. Although scenarios of no population change cannot be rejected for all individuals and for icteronotus, constant population size is firmly rejected for flammigerus, which shows strong evidence for demographic expansion

Fig. 6

Variation in morphology and coloration of males over c. 130 km across the Ramphocelus hybrid zone in historical and recent specimens. a-f Circles are values for individual specimens; dark lines are clines for traits and periods in which parameters were estimated using Hill functions, with shading indicating 95% confidence intervals around cline estimates. g, h Estimates of cline centers at each time period and their confidence limits estimated using bootstrap samples, revealing coincident eastward movement of the hybrid zone over time as indicated by both traits (colors as in a-f)

Fig. 7

Probability density distibutions of cline center and cline width parameters estimated across 1000 bootstrap samples of specimens for the 1911 (red), 1956 (blue), and 2010 (green) time periods. Both morphological PC1 and chroma show a shift in cline center and a reduction of cline width over time