The New World blackbirds (Icteridae) are among the most extensively studied avian clades in ecological, behavioral, and morphological research (Powell et al. 2014). This passerine family comprises approximately 106 extant species widely distributed throughout the Americas, from Alaska to Tierra del Fuego (Fraga 2008). Encompassing well-known groups such as New World orioles, grackles, cowbirds, and meadowlarks, icterids exhibit remarkable diversity in evolutionarily important traits, including plumage, vocalizations, social behavior, and foraging strategies (Jaramillo and Burke 1999). As a result, members of this clade occupy nearly all major ecosystems in the New World, ranging from boreal forests and temperate grasslands to tropical rainforests, savannas, shrublands, and wetlands (Winkler et al. 2020). It is therefore reasonable to expect that their ecological and phenotypic diversity is closely tied to their biogeographic history, with the origin of the clade and subsequent dispersal across the Americas driving distinct distributional patterns and adaptive specializations in response to diverse environmental pressures (Ricklefs 2006a; Simões et al. 2016). Nevertheless, despite the prominence of the clade as a model system in ecological studies, its biogeographic history remains both controversial and understudied.

In 1946, Ernst Mayr theorized that the current distribution of Icteridae resulted from continuous colonization from South America, even before the Isthmus of Panama closed (Mayr 1946). Thus, despite being Pan-American, Icteridae was established as a clade of Neotropical origin. This hypothesis remained widely accepted until a recent study by Barker et al. (2015), which placed Icteridae decidedly as a clade of North American origin. Although this study provided robust and compelling evidence for a North American origin of Icteridae, its broad approach prevented a more detailed understanding of the biogeographic history of the clade (Powell et al. 2014). As a result, key evolutionary processes, such as within-continental dispersal and in situ diversification, remain unresolved (Daru et al. 2017). For instance, it is still unclear whether the expansion of the group followed a stepwise pattern through adjacent regions, as Mayr originally proposed, or whether dispersal was more opportunistic and unconstrained, consistent with the high dispersal ability and migratory behavior of many icterid species (Kondo and Omland 2007; Arango et al. 2022). Similarly, the extent to which these dynamics and geographic barriers have shaped diversification across the Americas remains highly debated (Ricklefs 2006b; Machac 2020).

To better understand biodiversity patterns, it is crucial to define biogeographic boundaries that reflect historical processes (Olson et al. 2001; Kreft and Jetz 2010). Incorporating evolutionary information into boundary delimitation reveals both the spatial and evolutionary structure of species distributions (Daru et al. 2016, 2017; Kennedy et al. 2017) and sheds light on the mechanisms driving these patterns (Leprieur et al. 2012; Daru et al. 2017). Therefore, by identifying clusters of lineages that diversified within shared biogeographic contexts, evolutionarily informed bioregionalizations offer insights into spatial aspects of evolutionary history, such as dispersal and in situ diversification (Baselga and Orme 2012; Daru et al. 2017). This perspective is particularly important for clades such as Icteridae, which exhibit high ecological versatility and broad distributions across heterogeneous landscapes (Winkler et al. 2020).

As such, incorporating these evolutionarily informed bioregions into parametric models of range evolution can deepen the understanding of how geography and range evolution have shaped the diversification of lineages (Matzke 2013). This approach provides a more comprehensive view of the patterns, processes, and historical events that underlie present-day biodiversity distributions (Leprieur et al. 2012; Matzke 2014; Arango et al. 2025; Lima et al. 2025). Moreover, geography can also influence macroevolutionary dynamics by mediating speciation and extinction rates across regions that might differ in area, connectivity, ecological heterogeneity, and isolation (Goldberg et al. 2011; Rolland et al. 2014; Beaulieu and O’Meara 2016). In the case of Icteridae, North and South America could differ markedly in their biogeographic histories, climatic regimes, and habitat diversity, with differential diversification rates between the two regions further highlighting the main processes shaping the biodiversity of the clade (Ricklefs 2006b; Smith et al. 2014).

In this study, the historical biogeography of Icteridae is reconstructed, and the influence of geography on its diversification is assessed by integrating phylogenetic bioregionalization with comparative modeling. First, evolutionarily coherent bioregions across the New World are delineated using a Phylogenetic Bioregionalization framework (Daru et al. 2020), providing biologically meaningful units for ancestral range estimation with BioGeoBEARS (Matzke 2014). The most common models of range evolution are compared across three different hypotheses: an unconstrained model with equal dispersal among regions, a dispersal-restricted model incorporating geographic connectivity, and a directional model allowing asymmetric movement between North and South America. Biogeographic stochastic mapping (BSM) is then used to reconstruct the biogeographic events of the clade. To evaluate the impact of geography on diversification rates, Geographic Hidden State Speciation and Extinction (GeoHiSSE) models (Beaulieu and O’Meara 2016) are applied to estimate speciation and extinction across regions. Together, these approaches clarify how geography, dispersal, and regional dynamics have shaped the evolutionary history of Icteridae.

This study sheds new light on the historical biogeography of Icteridae, a widespread and ecologically diverse New World avian clade. By combining phylogenetic bioregionalization, probabilistic inference of historical range evolution, and geography-dependent diversification modeling, this study offers an integrated view of how biogeographic events shaped the evolutionary history of this family. The findings support a Nearctic origin for Icteridae followed by a progressive southward expansion and in situ diversification across the Americas, refining and expanding upon prior hypotheses about the evolutionary trajectory of the family and offering broader insights into the macroevolutionary mechanisms shaping New World avifauna.

The ancestral area reconstructions strongly support a Nearctic origin for Icteridae, corroborating the molecular phylogenetic evidence of Barker et al. (2015) and refuting earlier ideas of a Neotropical origin put forward by Mayr (1946). The reconstructed colonization sequence, from the Nearctic into Mesoamerica (~7.53 Ma), then to Amazonia (~6.06 Ma), the Andes (~5.84 Ma), Southern South America (~4.83 Ma), and finally Patagonia (~2.68 Ma), traces a stepwise southward progression consistent with the gradual DEC model that provided the best fit to the data. This pattern is consistent with broader trends identified across New World birds, where avian lineages with Nearctic origins have repeatedly dispersed into the tropics and radiated throughout South America, while lineages of South American tropical origin remain largely restricted to the Neotropical region (Smith and Klicka 2010). The asymmetry in dispersal directionality recovered for Icteridae, therefore, could reflect a broader macroevolutionary pattern of niche conservatism and southward ecological expansion that has shaped New World avifauna more generally (Hawkins et al. 2006; Smith and Klicka 2010; Smith et al. 2012).

Concordantly, most of the early speciation events in Icteridae occurred during the mid- to late Pliocene, coinciding with the progressive expansion of open habitats (Anderson 2006; Jaramillo 2023), which appears to be the ancestral habitat for the family (Suppl. material 1: fig. S5). This result is consistent with open habitat being the predominant habitat association among lineages at the base of the tree, including the meadowlark and allies clade (Xanthocephalus, Dolichonyx, Sturnella, and Leistes), which is sister to all other Icteridae. Further expansion of these habitats and their displacement of woody biomes as the planet became drier (Pagani et al. 1999; Osborne 2008) might have driven the expansion and dispersal of Icteridae, as its lineages were following their ancestral niche in open grass-dominated biomes (Wiens and Donoghue 2004; Crisp et al. 2009) or the retreating woody biome (Jaramillo 2023).

As such, the timing of the earliest dispersal from North America to South America (~6.06 Ma) predates the traditionally accepted establishment of the Isthmus of Panama (between ~3 and 2.7 Ma; Coates and Stallard 2013; Bourke et al. 2025), suggesting that early colonization of South America occurred via overwater dispersal or through pre- or proto-isthmian land connections. This timing agrees with the expansion of open habitats in northern South America due to Andean uplift approximately 9–6 Ma (Kirschner and Hoorn 2020). Moreover, the timing also coincides with growing biological evidence that the Great American Biotic Interchange (GABI) was a long and complex process beginning as early as the Oligocene–Miocene transition, with significant dispersal waves of terrestrial organisms documented at approximately 20 and 6 Ma (Bacon et al. 2015). While following the suggested pattern of northern colonization with subsequent speciation in South America (Smith and Klicka 2010), such events occurred at least 3 million years earlier than previously suggested, following more closely the timing of a dispersal wave (Bacon et al. 2015) than continuous dispersals after or during the establishment of the Isthmus ~3 Ma (Weir et al. 2009; Smith and Klicka 2010; Barker et al. 2015). This early dispersal of Icteridae to South America coincides with recent fossil evidence from North American mammals, which accumulated in Mesoamerica through the Miocene before a gradual exchange with South America approximately 7–3 Ma, also driven by the expansion of open habitats (Tseng et al. 2026).

Mesoamerica was both the first Neotropical region colonized from the Nearctic and the highest net source of dispersal events across the entire history of the clade, while simultaneously exhibiting the greatest number of range contractions and in situ speciation events among all regions. This combination of high dispersal, high speciation, and high lineage turnover suggests that Mesoamerica has acted as a highly dynamic biogeographic hub rather than a simple stepping-stone. As such, Mesoamerica harbors exceptionally high avian endemism generated by a complex geological and ecological history (Rocha-Méndez et al. 2019) and functions as a key relay in biotic interchange, where topographic heterogeneity promotes both diversification and lineage turnover (McCormack et al. 2023). The peak in range contraction counts in Mesoamerica approximately 3 Ma (Suppl. material 1: fig. S4) may also reflect competitive dynamics following land-bridge establishment, as newly arrived lineages displaced earlier colonists (Weir et al. 2009; Smith and Klicka 2010). At the same time, most in situ speciation events in the region occurred during the Pleistocene. Examples of these events include most Quiscalus and Icterus lineages, which had already begun to diverge during the Pliocene (Powell et al. 2014). The timing of these cladogenetic events may be related to rapid diversification driven by the characteristic climatic oscillations of this epoch, which caused extensive habitat fragmentation and ultimately allopatric speciation of already diverging lineages (Avise and Walker 1998; Avise et al. 1998; Weir 2006). In fact, such oscillations are well documented to isolate avian populations and generate the allopatric conditions necessary for speciation in Mesoamerica, particularly in open habitats (Barber and Klicka 2010; Castillo-Chora et al. 2021).

Within South America, the Amazonia region functioned as the primary continental entry point and dispersal hub to Southern South America even early on, with the second-highest mean in situ speciation after Mesoamerica. Its role as both a major source of dispersal events and a center of in situ speciation agrees with the general pattern documented across many other Neotropical clades, for which Amazonia is recognized as the primary source of Neotropical biodiversity, providing over 2,800 lineages to other regions (Antonelli et al. 2018). The Andes region seems to have functioned primarily as a dispersal corridor rather than a center of diversification, as it was both a source and recipient of comparable dispersal events to different South American regions (Suppl. material 1: table S8), contrasting with the well-documented role of the Andes as a diversification hotspot for other Neotropical avian clades, such as thraupids and New World sparrows, where topographic complexity and elevational barriers have promoted allopatric speciation (Hazzi et al. 2018). One possible explanation for this contrast is that, unlike montane-endemic radiations, avian lineages associated with open habitats, such as early icterids, have higher dispersal ability (Sheard et al. 2020) and thus may be more likely to use the Andes as a connective corridor than as an isolating barrier (Dauerman et al. 2026).

Colonization proceeded southward, with Amazonia contributing substantially to the assembly of the Southern South American fauna. In situ speciation events are relatively few in Southern South America, consistent with biogeographic patterns in other New World birds (Smith and Klicka 2010). With respect to Icteridae, Southern South America was a net recipient rather than a net source, likely reflecting the non-Neotropical origin of the clade. Interestingly, South American regions exhibited early in situ diversification during the Pliocene, coinciding with the firm establishment of open habitats (such as páramo, floodplains, and savannas) in South America (Kirschner and Hoorn 2020; Jaramillo 2023). Patagonia stands out across all biogeographic metrics as a very distinct region. It received dispersal almost exclusively from Southern South America, harbored virtually no in situ speciation, and showed the lowest range contraction counts. These features seem consistent with a region that might have been colonized late, rarely, and sparsely (Vuilleumier 1991; Bukowski et al. 2024). The nearly absent speciation signal is consistent with recent findings that colonization of new regions does not automatically generate elevated diversification (Kennedy et al. 2017; Arango et al. 2025).

Beyond the family-wide pattern described above, the major subclades of Icteridae differ notably in their individual biogeographic trajectories (see Suppl. material 1 for a full discussion). The meadowlarks and allies largely mirror the family-wide pattern, diversifying in North America during the late Miocene while retaining their ancestral open-habitat niche, with range expansion rather than range shift driving their spread into South America. The caciques, oropendolas, and allies show the opposite habitat trajectory, tracking retreating closed-forest habitats as they expanded from a North American origin into Mesoamerica and Amazonia, where speciation increased. The orioles diversified extensively within Mesoamerica during the Pliocene, with subsequent dispersal to South America during the Pleistocene. Furthermore, this group accounts for most reversals of the broader southward colonization trend of the family, with several South American lineages returning to North America (Winger et al. 2014). Finally, the grackles and allies split into two contrasting patterns: one lineage remained largely confined to Mesoamerica and North America until the Pleistocene, and a second radiated into Amazonia by the Miocene–Pliocene boundary alongside the expansion of open habitats (Jaramillo 2023). Together, these subclade-level differences suggest that the overall stepwise, unidirectional colonization pattern of Icteridae potentially conceals some heterogeneity in the timing, direction, and ecological correlates of range evolution among its major lineages.

Finally, the macroevolutionary analyses indicate that the full GeoHiSSE model provided the best fit, revealing that diversification dynamics in Icteridae are shaped by both observed geographic occupancy and unmeasured hidden states. This implies that geography alone does not fully capture the heterogeneity in speciation and extinction rates across the clade (Caetano et al. 2018).

The two hidden rate classes reveal distinct macroevolutionary regimes that are consistent with the biogeographic history recovered by the BSM analyses. In hidden rate class A, cladogenetic rates are moderate and geographically structured. NA endemics show the highest within-area speciation, consistent with the deep early diversification recovered in the Nearctic by the BSMs, while SA endemics show the lowest values with a non-negligible extinction fraction, pointing to a more dynamic macroevolutionary regime in the Neotropics, where speciation is offset by extinction, as documented across other South American passerine radiations (Ricklefs 2006b). As such, the elevated extinction fraction relative to its turnover among SA icterid endemics in hidden rate class A is consistent with the high range contraction rates recovered by the BSMs in Andean and Amazonian regions, suggesting that lineage-level turnover has been a persistent feature of the history of Icteridae in South America.

WS lineages in hidden rate class A show high cladogenetic rates, consistent with the ongoing intercontinental exchange documented in the BSMs and with the high dispersal ability and migratory behavior that characterize many icterid lineages (Kondo and Omland 2007; Price et al. 2009; Arango et al. 2022). This elevated rate mirrors the high within-area speciation observed in NA endemics under the same hidden rate class, suggesting that the underlying hidden state continues to promote cladogenesis in WS lineages even as they expand into South America. This pattern suggests that the state has a similar effect in North America alone and in lineages spanning North and South America and, as such, might reflect a process of biome tracking in which early icterid species dispersed into South America as open habitats expanded during the Pliocene (Anderson 2006; Kirschner and Hoorn 2020; Jaramillo 2023). This interpretation is consistent with the broader pattern of niche conservatism documented across icterid lineages (Eaton et al. 2008) and the timing of the transition of icterid lineages into South America (Fig. 1b).

In hidden rate class B, the relationship between geography and diversification is qualitatively different and decoupled from the pattern in class A. NA endemics show near-zero cladogenetic rates, while SA lineages show rates comparable to those of class A NA lineages and WS lineages show the highest cladogenetic rate of any state combination. Due to the parameterization, turnover in WS lineages reflects rapid cladogenesis across all three ranges at once rather than a conventional high-turnover cycle of speciation and extinction (Goldberg et al. 2011; Caetano et al. 2018). These lineages therefore split rapidly into endemic daughter lineages on each continent, a pattern that, at the intercontinental scale, mirrors the dominance of cladogenetic range splitting documented across emberizoid passerines (Arango et al. 2025).

This contrast in cladogenetic rates depending on hidden states demonstrates that geography does not drive diversification in isolation. Instead, the same geographic area yields distinct diversification rates depending on the hidden regime of the lineage, indicating that geography merely contributes to a more complex rate structure. Importantly, the near-zero subsequent diversification of North American endemics in hidden rate class B does not reflect a geographic constraint on diversification in North America per se, as the Nearctic clearly supported substantial diversification in rate class A. This pattern indicates that lineages in hidden rate class B that happen to become NA endemics do not subsequently radiate (Caetano et al. 2018).

As such, mean speciation rates were comparable between continents, suggesting that both regions contributed substantially to overall clade richness, while the rate of lineage splitting in Icteridae is predominantly governed by an unmeasured biological axis that does not covary with either continent or range size. This finding resonates strongly with recent work on the broader Emberizoidea superfamily, which showed that broad geographic dispersal is not a diversification driver for the group and that cladogenetic events occur primarily within bioregions rather than between them, suggesting that ecological opportunity, habitat specialization, and interspecific relationships within cohesive bioregions drive speciation more than dispersal events between regions (Arango et al. 2025). Specifically, ecological specialization has been identified as a significant positive predictor of speciation rates across Emberizoidea at continental scales, including in Icteridae (Conway and Olsen 2019), while dispersal ability shows no significant relationship with speciation rates in the clade (Arango et al. 2026). Moreover, sexual selection through plumage and song divergence is recognized as a critical process for speciation (Higashi et al. 1999). In Icteridae and allied groups in particular, color evolution and dichromatism have been shown to potentially drive species cladogenesis (Friedman et al. 2009; Price and Eaton 2014). These lines of evidence point toward ecologically driven within-region speciation that could be mediated by feeding specialization, social behavior, and sexual selection as the most likely biological axes underlying the hidden diversification regimes recovered by GeoHiSSE, rather than any geographically structured process. Formally testing this hypothesis will require integrating these traits directly into future diversification analyses of Icteridae.

Taken together, these results suggest that the diversification history of Icteridae was shaped by the interplay between geography-mediated colonization, regional ecological opportunity, and different macroevolutionary regimes, with the Nearctic providing an ancient source of deep early diversification (Arango et al. 2025); Mesoamerica acting as a dynamic filter, relay, and site of Plio-Pleistocene in situ diversification (Barber and Klicka 2010; Castillo-Chora et al. 2021); and the core Neotropical regions accumulating diversity through a combination of immigration and in situ speciation (Smith and Klicka 2010; Harvey et al. 2020). Meanwhile, the macroevolutionary results establish that speciation rate is primarily governed by factors other than geography, with a secondary geographic component that is consistent with—but does not strongly drive—the continental patterns recovered by the BSMs. These results not only corroborate the origin of Icteridae posited by Barker et al. (2015) but also support the stepwise colonization hypothesis of Mayr (1946), although the findings indicate that colonization occurred in the reverse direction.