Species are the fundamental unit for measuring and conserving biodiversity. Delimiting species and mapping their distributions are the first steps toward their conservation. However, we are still far from fully understanding how many species are on Earth, where they occur, and the mechanisms that shape this diversity and its biogeographic patterns (Lomolino 2004; Whittaker et al. 2005).

A striking and intriguing pattern in the biogeography of species is the disjunct distribution, where otherwise closely related populations are separated by large distances, even with no evident geographical or ecological barriers. Sampling biases can be behind an apparent but not real disjunct distribution, which is perceived as such due to incomplete sampling in intermediate areas or species having low detectability (Taylor 2017). Once these artifacts are tested and discarded, a clustered geographical pattern may be due to three processes: vicariance, local extinction, or long-distance dispersal events (Taylor 2017; Owen and Taylor 2023). Natural vicariance (i.e., not due to human intervention through habitat fragmentation and other barriers of anthropogenic origin) may be caused by physical barriers and historical climatic changes that create unsuitable habitats (e.g., Kropf et al. 2006; Pârvulescu et al. 2019), potentially extirpating populations and dividing the species’ once-continuous range into geographical clusters. Local extinction may result from historical changes in ecological conditions, either abiotic or biotic, leading to the extirpation of populations from otherwise apparently current suitable habitats (Harrison 1991). Finally, long-distance dispersal events can also lead to disjunctions when jump dispersal processes establish new populations in suitable habitats that are geographically isolated from the main distribution range (Marin da Fonte et al. 2019).

The three mentioned drivers can promote allopatric speciation if the spatial or temporal gap between population clusters sufficiently restricts gene flow. Indeed, molecular data have frequently revealed that disjunct populations previously considered conspecific actually represent distinct species (e.g., Carranza and Amat 2005), especially when such populations had been at some point previously recognized as subspecies (e.g., Fok et al. 2002). Thus, assessing the specific identity of clusters within a disjunct-distributed taxon can lead to the discovery and description of new species or confirm such disjunct distribution.

Amphibians constitute valuable, yet little studied models for biogeographic research: most species are small and tightly tied to humid environments, which limits their dispersal and makes their distributions especially informative of environmental constraints, while some taxa retain enough vagility to colonize new areas, offering insight into range-expansion dynamics. Our knowledge of the taxonomy and distributions of amphibians—the most endangered class of vertebrates (Luedtke et al. 2023)—, has improved significantly in recent times, partly due to the integration of molecular data into systematic studies (e.g., Veith 1996; Frost et al. 2006; Pyron and Wiens 2011). This has enabled the description of ca. 100 new species of frogs per year in the last decades (Köhler et al. 2005; Frost 2026) and the resolution of numerous species complexes, thereby improving our understanding of species ranges (Streicher et al. 2020; Frost 2026). However, despite these recent advances, knowledge of amphibian distributions remains fragmentary in many regions (Brito 2010), particularly in tropical and sub-tropical areas —the most biodiverse and imperiled on Earth (Barlow et al. 2018). This is largely caused by three reasons: (1) Even in the best-studied tropical biodiversity hotspots, species inventories remain incomplete and taxonomic uncertainty continues to obscure what is actually being mapped (Carné and Vieites 2024); (2) amphibians are typically small, secretive, and often exhibit seasonal activity patterns, which reduces their detectability —particularly in strongly seasonal habitats where active periods are short (Dubos et al. 2020); and (3) field trips are logistically complex, costly, and time-consuming, leading surveys to focus on the most accessible or attractive-to-sample sites (Sastre and Lobo 2009), while less appealing habitats (e.g., plantations) are often overlooked —despite their potential to provide a more complete picture of species’ distributions and ecologies (e.g., Carné et al. 2025). Furthermore, DNA barcoding is not yet widely implemented (Carné et al. 2026), creating uncertainty when assigning localities to certain species based solely on morphology. This is especially problematic when closely related species exhibit intraspecific morphological variability or phenotypic conservatism in relation to their congeners.

The anuran genus Pleurodema Tschudi, 1838 comprises 15 described species of leptodactylid frogs distributed across the Neotropics (used here in its classical sense; but see Holt et al. 2013), from Panama to southern Argentina and Chile, and occurring from sea level up to 5,000 m a.s.l. (Faivovich et al. 2012; Frost 2026). Most species primarily inhabit open environments, often in arid and/or cold, or highly seasonal temperate regions (Ferraro and Casagranda 2009; Faivovich et al. 2012). Due to these harsh conditions, the breeding season is short, with optimal mating and egg-laying opportunities sometimes limited to just a few days per year, thus leading most species to exhibit explosive breeding behavior (Hödl 1992; Ferraro and Casagranda 2009; Faivovich et al. 2012).

Several studies have been carried out on the phylogeny of Pleurodema (Duellman and Veloso 1977; Faivovich et al. 2012; Ferraro 2022). All of them recover a clade containing two sister, morphologically similar species from northwestern Argentina, Pleurodema nebulosum (Burmeister 1861) and P. guayapae Barrio, 1964. Pleurodema guayapae was distinguished from P. nebulosum based on clear bioacoustic differences and the absence of an external tympanum; hybridization near the type locality was documented (Barrio 1964). These two species have been hypothesized to have diverged during the Pleistocene due to barriers such as rivers carrying glacial runoff or the uplift of N–S mountain ranges (Duellman and Veloso 1977), but this has not been corroborated.

Pleurodema guayapae (Fig. 1) is a species that typically inhabits saline flat environments and semidesert regions of the Dry Chaco ecoregion, and was originally known from the convergence zone of the Argentine provinces of Catamarca, Córdoba, La Rioja, San Juan, San Luis, and Santiago del Estero (Cei 1980; Guerreiro et al. 2005; Ferraro and Casagranda 2009; Quiroga and Sanabria 2009). Surprisingly, De la Riva and Gonzales (1998) reported morphologically identified P. guayapae specimens from four nearby localities in anthropogenic habitats of Bolivia’s Santa Cruz department, approximately 1000 km north of the nearest Argentine population known at that time. Subsequent surveys since 1998 have expanded the species’ known Argentine distribution northward but, despite extensive sampling efforts in intervening areas, a significant 820 km distributional gap persists between the northernmost Argentine records (Santiago del Estero Province) and the Bolivian populations. The unexpected discovery of P. guayapae in Bolivia raised questions concerning the taxonomic status of Bolivian populations (De la Riva and Gonzales 1998; De la Riva et al. 2000; Faivovich et al. 2012). If Argentine and Bolivian populations were conspecific, we could consider P. guayapae a proven example of confirmed disjunct distribution in amphibians.

In this study, we use molecular data to evaluate —twenty-seven years after their discovery— the taxonomic identity of the Bolivian populations morphologically attributed to P. guayapae. Our main goal was to determine whether the Bolivian specimens are conspecific with those from Argentina and thus confirm a hypothesized bicentric species distribution (i.e., geographically separated but genetically undifferentiated populations), or whether they constitute a distinct vicariant taxon (i.e., geographically and genetically differentiated). Additionally, we explore the potential ecological and historical factors that might have shaped the biogeographic patterns.

Despite the substantial ~ 820 km geographic gap between the two known population clusters of P. guayapae —an area lacking occurrence records of the species even after decades of field surveys (see Suppl. material 4)— Bolivian and Argentine populations show no significant genetic divergence in the mitochondrial markers analyzed (16S and Cyt-b) and do not form reciprocally monophyletic clades. These findings suggest that the observed disjunct distribution likely reflects a relative recent process rather than prolonged isolation. The fact that these populations do not represent genetically distinct lineages at mitochondrial level primarily supports the hypothesized true disjunct (i.e., bicentric) distribution of a single species. Nonetheless, data from nuclear genes are needed to test whether population structure or isolation exists but is obscured by mitochondrial introgression.

The most intuitive explanation would be to attribute the disjunct distribution to a sampling bias. Many Pleurodema species, among them P. guayapae, spend much of the year buried, which hampers detectability outside of explosive breeding events that occur during brief and unpredictable windows of optimal environmental conditions (De la Riva and Gonzales 1998; Valetti et al. 2014). It would seem therefore plausible that undetected populations might occur in the intermediate areas. However, this appears unlikely because the regions between the known population clusters of P. guayapae have been relatively well surveyed from a herpetological standpoint (e.g., Lavilla and Scrocchi 1991; Cruz et al. 1992), with survey intensity (i.e., number of reptile and amphibian specimen records) comparable to, or greater than, that of the localities where the species has actually been recorded (see Suppl. material 4 for distribution and sampling intensity maps of amphibian and reptile records within the study area). Furthermore, despite the species’ secretive habits, Schalk and Saenz (2016) reported calling activity from January to May and again in October —thus substantially extending the potential detection window. Moreover, P. guayapae lays eggs in floating foam nests and has exotrophic tadpoles that develop in lentic water bodies (Valetti et al. 2014), both of which are conspicuous and increase the likelihood of detection. These factors strongly suggest that the species is genuinely absent from the intervening region, and thus we can assume that the mentioned gap is real and not a human artifact due to lack of surveys. However, in order to confirm this hypothesis, we recommend that future herpetological surveys in the intervening area put a special focus on the species by searching in appropriate environments, especially during explosive breeding events of other chacoan amphibians, and by checking the identity of the tadpoles found at temporal water bodies.

Once the sampling bias is discarded, the disjunct distribution pattern of P. guayapae may have two alternative explanations: long recent dispersal, or a complex history shaped by climate changes and including local extinctions.

The first hypothesis is that the observed distribution results from a recent long-distance dispersal event. Amphibians are generally poor long-distance dispersers (Duellman and Trueb 1986) due to small body size, physiological constraints, habitat specificity, and phylopatry, making such events rare and typically passive (Marin da Fonte et al. 2019; Owen and Taylor 2023). If such a dispersal occurred, it was likely human-mediated. This hypothesis gains some support from the fact that all known Bolivian populations of P. guayapae occur in anthropogenic habitats (De la Riva and Gonzales 1998), suggesting that human-altered landscapes may have facilitated the passive transport and establishment of individuals—possibly through the movement of soil or agricultural materials where frogs could have been concealed. The fact that most Bolivian localities of P. guayapae are along the railway connecting Bolivia and Argentina might support this possibility; however, the trains do not continue southwards from the Bolivian border, and from that point (Yacuiba) there is still a gap > 550 km airline to the closest P. guayapae localities in Santiago del Estero. Thus, the plausibility of a human introduction is limited, as transporting material from rural areas over more than 800 km with no direct connections and across international borders is logistically unlikely.

According to our results, the most plausible explanation for the current disjunct distribution is that past climate fluctuations triggered multiple episodes of range fragmentation, contraction, and expansion since the LGM (Fig. 2). During the Heinrich Stadial and Bølling-Allerød periods, several isolated but climatically suitable habitat clusters emerged (see Results). However, declines in habitat suitability in subsequent periods suggest that any populations occupying many of these clusters may have undergone local extinction. Throughout all periods, the most suitable habitat cluster occurred near the borders between Argentina, Paraguay, and Bolivia, suggesting that this region may have acted as a source cluster for the species’ present distribution. Habitat suitability patterns during the Northgrippian (8.3–4.2 ka) further suggest that the formerly monocentric core became fragmented in multiple clusters, and that this fragmentation, combined with subsequent range contractions to the centers of these clusters ultimately shaped the current disjunct distribution.

In our area of study, some investigations have dealt with disjunct distributions shown by populations of non-chacoan species now occupying different biogeographic domains or main ecosystems, for which the Pleistocene climatic fluctuations meant distributional shifts across what is today the chacoan realm—for example, the harvestmen Geraeocormobius sylvarum and Discocyrtus dilatatus (Acosta et al. 2007; Vergara et al. 2017). Likewise, many chacoan organisms experienced distributional shifts during those climatic fluctuations, as the ecosystem moved northwards during cool periods and recovered during warm ones, causing range expansions and contractions and leading to the disjunct or patchy distributional patterns observed today. Some examples include certain leguminous plants (Robbiati et al. 2021; Sinani et al. 2022), palms (Silva et al. 2023), and mammals (Ferreiro et al. 2022). Among amphibians, the most striking case of disjunct distribution is that of P. guayapae reported herein. However, it is not an isolated case: within the same study area, Chacophrys pierottii shows a similar pattern that warrants further investigation (IUCN SSC Amphibian Specialist Group 2021a), and so does Lepidobatrachus llanensis (IUCN SSC Amphibian Specialist Group 2021b), whose disjunct populations have already been genetically confirmed to be conspecific (Faivovich et al. 2012).

The Gran Chaco ecoregion, which expands over more than one million square kilometers across Argentina, Bolivia, Brazil, and Paraguay, is suffering a catastrophic rate of deforestation and alteration, mostly due to the expansion of intensive agriculture, cattle ranching, and fires (see Mosciaro et al. 2022, and references therein). Altered habitats can act as ecological filters, modulating the response of amphibian communities to land-use conversion in a way that does not necessarily follow a simple richness-decline pattern (Assmann et al. 2026). Some species of amphibians like P. guayapae seem to tolerate a certain degree of human perturbation, and likely have not been as drastically affected by land use changes as others (see Medina et al. 2016). In the future, the systematic and rapid destruction of this rich ecosystem might extirpate many organisms characteristic of it, including species that remain poorly studied, thus precluding the understanding of their distributional patterns and the past processes that shaped them. In such an undesirable scenario, questions as intriguing as the disjunct distribution of P. guayapae might never be addressed.

Twenty-seven years after the discovery of the Bolivian populations of Pleurodema guayapae, we demonstrate that, despite the ca. 800 km separating the two geographic clusters and the absence of records from the intervening areas notwithstanding extensive sampling efforts, the Bolivian populations are conspecific with those from Argentina. This represents a striking example of a true disjunct distribution in amphibians, likely resulting from repeated range expansions and contractions driven by past climatic oscillations.