Las inferencias filogeográficas, dentro de un marco comparativo cuando incluyen varias especies a escala regional, permiten detectar patrones de biodiversidad regional y de paisaje (fisonómicos), importantes para entender tanto la macroecología como el impacto a gran escala de los eventos geológicos. En los últimos años, el conocimiento filogeográfico de Patagonia se ha acumulado para organismos acuáticos y terrestres, y aunque se han propuesto pocos patrones demográficos o espaciales cualitativamente concordantes, no se ha hecho ningún intento de revisar comparativamente algunos de esos patrones considerando grandes grupos de organismos. En esta revisión compilamos el conocimiento publicado sobre la filogeografía de plantas vasculares y vertebrados terrestres de Patagonia con el propósito de comparar niveles de diversidad genética, procesos demográficos, quiebres filogeográficos y localización de posibles refugios, para producir el primer resumen de patrones filogeográficos emergentes para esta región. Esta revisión revela algunos patrones filogeográficos congruentes dentro y entre plantas vasculares y vertebrados terrestres, y sugiere que tanto los eventos geológicos Cuaternarios como los pre-Cuaternarios pudieron haber jugado un papel importante en la historia evolutiva de los linajes patagónicos. Diferentes procesos y cambios direccionales en los rangos de distribución, sugieren un mosaico de patrones filogeográficos mucho más complejo que los patrones norte-sur tradicionalmente propuestos.
Phylogeography is an integrative field of science linking micro- and macroevolutionary processes (Bermingham & Moritz, 1998). By combining analyses of current population genetic structure with organismal distributions and historical changes in the environment, phylogeography provides the means for making strong inferences regarding dispersal, speciation, extinction, and other population-level processes. Applied in a comparative framework across multiple species at a regional scale, phylogeographical inference enables the detection of regional and landscape-level patterns of biodiversity important for understanding macroecology, broad impacts of geological events, and areas of high conservation priority (Bermingham & Moritz, 1998; Moritz, 2002).
Among inhabited continents, South America exhibits the greatest void in phylogeographical knowledge, with only 6.3% of relevant articles published worldwide from 1987 to 2007 dedicated to organisms of this continent (Beheregaray, 2008). Unlike temperate regions of North America and Europe, where considerable effort has focused on phylogeny reconstruction within and among species, very little is known about the phylogeography of temperate South America. Patagonia, the southernmost area of this continent, is among the least studied regions phylogeographically, and yet arguably has been impacted more than other areas of South America by environmental changes, both recent and over geological time scales (Rabassa, 2008a).
Over the past several million years, glaciations and associated climate change, tectonics with volcanism, palaeobasins, and seashore shifts with marine introgressions have dramatically altered the landscape of Patagonia (Ramos & Ghiglione, 2008; Martínez & Kutscher, 2011; Ponce et al., 2011). The numerous glacial advances and retreats during the Pleistocene are also well known (Rabassa, Coronato & Martínez, 2011). For example, the Greatest Patagonian Glaciation (GPG; 1.2 to 1.0 Mya) and the Last Glacial Maximum (LGM; 20 to 18 kya) differed both in extent and duration. The ice sheet extent during the GPG was significantly greater than that during the LGM and the last several glaciations (McCulloch et al., 2000; Ponce et al., 2011); thus, these events most likely differentially impacted the abundance and distribution of the local biota.
Importantly, shared biogeographical patterns alone are insufficient for inferring causal mechanisms. Similar concordant patterns may result from different underlying causal factors operating over different times (i.e. pseudocongruence; Soltis et al., 2006). However, if diverse taxa with variable natural histories and surveys based on different kinds of genetic markers produce qualitatively congruent phylogeographical patterns, then it is likely that the geological and environmental history of a certain region produced similar impacts on the regional biota. Building upon single-species analyses, the first step in comparative phylogeography is to determine whether qualitatively congruent patterns exist among taxa that can serve as initial hypotheses. These hypotheses can further be tested with independent data, new analyses, and rigorous comparative statistical frameworks (Knowles, 2004; Carstens & Richards, 2007; Knowles & Carstens, 2007; Hickerson et al., 2010).
The goal of the present study is to review and summarize current knowledge about the phylogeography of plants and terrestrial vertebrates from Patagonia. We compare levels of genetic diversity, inferred demographic processes, phylogeographical breaks and putative refugia to produce the first regional summary of emerging phylogeographical patterns and past scenarios for this region.
MATERIAL AND METHODS
Patagonia is defined here as the region extending south of the Río Colorado (approximately 35°S to 36°S) in Argentina, and in Chile, south of Valdivia (approximately 39°S) to the southernmost point of South America (56°S). This area includes two main geographically distinct units. First, the Patagonian steppe is a large, dry, extra-Andean plain covered by grassland and scrubby vegetation that extends from the eastern slopes of the southern Andes to the Atlantic coast, and a small portion of Chile (between 36°S and 43°S). This extension includes two main phytogeographical provinces: the Monte Province which extends from central-northern Argentina to north-eastern extra-Andean Patagonia, reaching 43°S and the Patagonian Province which stretches over the remaining extra-Andean extensions. The second unit, the Andean region, is much smaller and runs along the Andes from 37°S to the southernmost Magellan and Fuegian Archipelagos. This unit contains two main phytogeographical provinces: the Sub-Antarctic province which comprises different forest types on the eastern and western slopes of the Andes and several units of grassy steppes, and the Alto-Andean province which occupies the highest mountains between 1600 and 600 m a.s.l. (Coronato et al., 2008).
We compiled all published phylogeographical studies of Patagonian plants and terrestrial vertebrates. A few other studies, not specifically of a phylogeographical nature (e.g. isozymes studies) but that also provide information on genetic diversity in a geographical and historical context, were also included. We searched in the Web of Science for all articles that contained in the title, abstract, or keyword the following terms: ‘Patagonia phylogeography’, ‘Patagonia Pleistocene glaciations’, ‘glacial refugia Patagonia’, and ‘South America biogeography’.
From each paper, we recorded the following information: geographical coordinates, number of sampled sites (localities), number and type of molecular markers, geographical extension, species range coverage, type and location of phylogeographical breaks, hypothesized refugia, secondary contact areas, inferred demographic processes by area and divergence times (if available).
To summarize the sampling locality coverage, all localities were mapped using the point-to-grid richness analysis tool in DIVA-GIS (Hijmans et al., 2001), using a 0.1667 × 0.1667° grid (i.e. 18 × 18 km at the equator), and applying the circular neighborhood option with a 1° (111 km) diameter. We made two maps (plants and vertebrates) from the complete dataset; and four additional maps were made to focus solely on lizards, rodents, woody plants, and perennial herbs.
Refugia and breaks
We recorded the type and geographical distribution of hypothesized refugia (i.e. in situ persistence during Pleistocene glaciations) for each species. Some of these were explicitly mentioned as such in the original papers, whereas others were more subtly mentioned, mainly in the discussion sections. Generally, studies have recognized refugia based at least on one (and often on several) of the following criteria: restricted areas with the presence of exclusive haplotypes/alleles, shared haplotypes /alleles between potential refugia and recolonized areas, and high genetic diversity. Based on their geographical localization we classified glacial refugia into the following types: peripheral and lowland (sensuHolderegger & Thiel-Egenter, 2009) and valley refugia (i.e. patchy refugia produced by Alpine valley Glaciations; Rabassa & Clapperton, 1990). We also compiled information on geographical genetic discontinuities to summarize phylogenetic breaks.
Geographical distributions of genetic diversity were explored within each principal organismal group (i.e. plants and vertebrates) and in combination. Diversity indices per population (in plants) or clade (in vertebrates) were standardized and scaled from 1–100 within each species. We selected whether the haplotype diversity index or its equivalent for diploid data, the expected heterozygosity. In studies where diversity indices were not shown for each population or clade but information was available, we calculated gene diversity following the formula proposed in the ARLEQUIN software manual (p. 90; Excoffier, Laval & Schneider, 2005). Estimated genetic diversity for each group of organism (i.e. plants and vertebrates) was mapped using the point-to-grid statistical analysis tool in DIVA-GIS (Hijmans et al., 2001), using the mean as the output variable. To avoid misinterpretations of the data, two different grid sizes of 0.1667 × 0.1667° and 1 × 1° (i.e. 18 × 18 and 111 × 111 km at the equator, respectively) and the circular neighborhood option with a 1° (111 km) diameter were applied. To determine shared areas of high, medium, and low diversity between plants and vertebrates, each grid was reclassified in three levels (dividing the observed values range in three equal intervals: low, medium, and high genetic diversity) and then multiplied using the option overlay in DIVA-GIS. As an alternative approach, each species grid was reclassified in three levels, dividing the observed values range in three equal intervals: low, medium, and high genetic diversity; for each group of organisms, we performed a stack of all the reclassified grid layers, and the genetic diversity for each group was mapped using the point-to-grid statistical analysis.
For information on events and processes, we recorded the population geographical coordinates associated with the inferred process. Processes inferred by each reviewed paper were compiled either from nested clade phylogenetic analyses (Templeton, Routman & Phillips, 1995; Templeton, 2004), mismatch distribution (Rogers, 1995), and neutrality tests (Tajima, 1989; Fu, 1997). We also considered inferences that were based on analyses of the geographical distribution of genetic diversity. We grouped inferences into four categories: demographic stability (extended isolated areas where presumably large panmictic populations persisted), demographic/range expansion (including long distance dispersion/colonization), secondary contact, and fragmentation (including allopatric fragmentation and restricted gene flow) defined as an area containing several spatially circumscribed haplogroups being differentiated by relatively large mutational steps under allopatric fragmentation; we considered also colonization routes that were suggested in the reviews based on the distribution of genetic diversity from putative source areas (with high genetic diversity) to areas of low diversity, and/or the presence of shared versus private haplotypes. For each group of organisms, populations with their associated inferences were mapped using DIVA-GIS. Polygons indicating different processes were traced using CorelDraw (Corel Corp.).
A total of 33 publications were identified from the database search, 20 of which corresponded to plant studies and 13 to terrestrial vertebrates. Almost all plant studies (90%) focused on a single species, whereas 53% of vertebrate publications included more than one taxon. On average, plant studies included considerably higher number of individuals than vertebrate studies, whereas the number of sampling localities is similar for both (Table 1). Studies performed on vertebrates are methodologically more homogeneous than the plant studies. The vertebrate studies all used mitochondrial DNA as the molecular marker whereas, in plants, isozymes followed by cpDNA were the two most common molecular markers (Table 1).
Table 1. Summary statistics over 33 publications referred to phylogeographical studies on plants and terrestrial vertebrates from Patagonia
Species or complexes
Populations/species or complexes, mean (range)
Individuals/species or complexes, mean (range)
535 (60–2 608)
566 (60–2 608)
cpDNA or mtDNA
Total number of species studied with at least one marker.
Mean number of markers studied per species for the whole data set.
We tallied 1361 sampling localities (639 corresponding to terrestrial vertebrates and 722 to plants; Table 1). Sampling coverage overall species extended over the entire Patagonian region; however, sampling intensity varied when considering terrestrial vertebrates and plants separately (Fig. 1). Terrestrial vertebrates were more intensively and homogeneously sampled at mid latitudes (between 34° and 44°S), covering the steppe as well as the Andean regions (Fig. 1A). Within vertebrates, lizards were homogeneously sampled, except for the southernmost part of Patagonia where sampling was less intense (Fig. 1B). Rodents were not only more broadly, but also more heterogeneously sampled, with a partial gap on the Santa Cruz steppe (Fig. 1C).
Plants, on the other hand, were more heterogeneously sampled than vertebrates. They were most densely studied in the Andean region between 38° and 45°S, whereas the steppes of Río Negro and Chubut Provinces are almost unexplored (Fig. 1D). Considering woody plants and perennial herbs separately, woody species were exclusively sampled on the Andean forests, where they have been intensively studied around latitude 41°S (Fig. 1E). Perennial herbs were more heterogeneously sampled, with intensive sampling along the Andean range but only limited sampling on the steppe (Fig. 1F).
All reviewed studies on terrestrial vertebrates covered the entire (or nearly entire) extension of the known species geographical range. By contrast, only 48% of the plant studies covered the total distributional range of the focal species.
We were able to extract information on putative refugial areas for both vertebrates and plants (Fig. 2 and Table 2). Although fewer in number for vertebrates (Fig. 2A), putative refugia were highly coincident with putative refugia proposed for plants (Fig. 2B). Overall, five refugial areas were distinguished, which we grouped into three categories (Fig. 2): ‘lowland refugia’, in the Chilean-Coastal range north of 42°S (a), in the coast of the San Jorge Gulf (approximately 46°S; b) and in southern Patagonia (approximately 51°S to 53°S; c); ‘peripheral refugia’, a longitudinal zone located along the eastern and western flanks of the Andes between 36°S and 53°S (d); and ‘valley refugia’, in the high-Andean region north of 36°S (e).
Table 2. Location and type of putative glacial refugia proposed in the reviewed articles
For terrestrial vertebrates, two phylogeographical breaks have been reported in the southern Central Andes at 33°S (break 1, 1 sp.) and 35°S (break 2, 1 sp.), and a third one for the Coast of Central Chile at 38°S (break 3, 4 spp.). For the steppe area, six breaks were detected, four of them only for one species each, following approximately different river basins: Río Colorado (break 4), Río Negro (break 5), Río Limay (break 6), and the Río Chubut (break 7; Fig. 2A). The other two were concordant for two species each, one coincident with the Somuncurá plateau (break 8, 42°S), and the second located at 45°S (break 9, northern limit of the San Jorge Gulf). The southernmost break, coincident with the Magellan strait, was recorded for only one species (break 10; Fig. 2A).
For plants, breaks have been reported along the Patagonian Andes at approximately the following latitudes: 35°S (break 2, 1 sp.), 37.5°S (break 3, 3 spp.), 40.5°S (break 4, 5 spp.), 43°S (break 5, 5 spp.), and 50°S (break 6, 1 sp.). Across the Patagonian steppe, three breaks were detected for one species each, following approximately different river basins, from north to south: Río Chubut (break 7), Río Chico (break 8), and Río Coig (break 9). The southernmost break, detected for one species, was coincident with the Strait of Magellan (break 10; Fig. 2B).
Four phylogeographical breaks are shared between some vertebrate and plant taxa. Two are located in the Central Andean Region approximately at the same latitude (35°S and 37.5°S; breaks 2 and 3; Fig. 2), one at the Strait of Magellan (break 10) and another along the Chubut river basin (break 7; Fig. 2).
Of the 20 plant studies, 17 (encompassing 640 sampling localities) included information regarding geographical patterns of genetic diversity. Diversity indices varied among studies with the two most common being expected heterozygosity and allelic richness. In vertebrates, 12 of the 13 studies (412 sampling sites) reported haplotype and nucleotide diversity indices for species or clades within them (see Appendix, Table A1).
In vertebrates, there were four main regions with high diversity (Fig. 3A): Chiloé Island, West-Chilean coast (north of 38°S), northern Neuquén and southern Mendoza Province in Argentina (35°S to 37°S) continuing southeast, following the Monte-Patagonia ecotone region, and the coastline of Rio Negro and Chubut Provinces in Argentina. Regions of low diversity were located along the Andean cordillera over almost the entire north–south axis of Patagonia (approximately 41°S to 52°S) in previously glaciated areas.
In plants, there were seven areas of high diversity (Fig. 3B): the West-Chilean coast (approximately 38°S to 42°S), southern Central Andes (approximately 35°S), northern Patagonian Andes (approximately 46°S), four areas covering all of the latitudinal extension of Patagonia located along a longitudinal transect of approximately 70°S (approximately 32°S, 39°S, 43°S, 49°S), central Patagonian steppe near the San Jorge Gulf (approximately 45°S to 46°S), southern San Jorge Gulf (approximately 47°S), and southernmost Patagonia and Tierra del Fuego (approximately 51°S to 53°S). Along the Atlantic coast, cells with intermediate values of genetic diversity predominate, whereas, towards the north-eastern region, cells with low diversity values are abundant. The regions of lowest diversity were located along the Andes cordillera over almost the entire foothill region of the southern Patagonian Andes (approximately 46°S to 52°S) in previously glaciated areas.
For plants and vertebrates combined (Fig. 3C), five regions with high levels of diversity were identified: Chiloé Island, in the West-Chilean coast (approximately 40°S), in the southern Central Andes (approximately 35°S), in southwestern Chubut Province in the Patagonian steppe (approximately 46°S), and southern San Jorge Gulf (approximately 47°S). Three additional areas showed moderate to high levels of genetic diversity: one located in the West-Chilean coast (approximately 37°S, 39°S, and 42°S), in the northern Patagonian Andes (approximately 35°S to 38°S), and in Central Patagonia (approximately 45°S to 47°S) ranging from the eastern Andean flanks to the center of the Patagonian steppe. Three regions of low diversity were also identified: the Atlantic Coast (approximately 47°S), southern Patagonian Andes (approximately 46°S to 53°S), and the northern Patagonian Andes (approximately 37°S to 43°S). Areas where plants and vertebrates presented contrasting diversity values (high versus low) were also detected, with the most prominent region located in the transition zone between Patagonia and the Monte phytogeographical regions where animals seem to have high and plants low diversity (Fig. 3C).
For vertebrate taxa, five areas were identified associated with fragmentation inferences (Fig. 4A and Table 3): (1) northern Neuquén province, (2) southern Neuquén and northern Chubut provinces, (3) Sotomayor High Plateau in the eastern coastline of Chubut Province, (4) a strip west of the San Jorge Gulf, and (5) the Strait of Magellan. Stable areas as defined here are located in (I) Chile north of latitude approximately 36°S (II) at approximately 41°S and (III) north of Chiloé Island. In Argentina, stable areas within Patagonia are located on both sides of the Río Colorado (IV, southern Mendoza and northern Neuquén), (V) northern Rio Negro, (VI) along a narrow east–west strip in northern Chubut, (VII) on Chubut northern coastline, and (VIII) in Tierra del Fuego (Fig. 4A). Several expansion routes have been inferred, some of them from the southern Chilean stable area southeast into Argentina and, in some rodent cases, reaching the southern tip of Santa Cruz Province. Southern expansion routes also have been proposed from the east towards the southwest at different latitudes. A third principal expansion has been proposed from northern Neuquén towards the south. Additional expansion routes, although with a south–north direction from southern and northern Chubut and northern Rio Negro, have also been proposed (Fig. 4A). Some of these routes could be associated with the detected secondary contact areas on (a) the northern tip of the San Jorge Gulf, (b) on the coastline of Rio Negro Province, (c) in the central area of the Somuncurá Plateau, and (d) in the ecotone area between the Patagonia and Monte eco-regions (Fig. 4A).
Table 3. Demographic inferences reported for each taxon
RE, range expansion; SC, secondary contact; FRA, fragmentation; S, stable areas; CR, colonization routes.
In plants, three geographical regions affected by fragmentation processes were identified (Fig. 4B and Table 3): (1) central-eastern Patagonia, (2) Andean flanks north of 38°S, and (3) the north-eastern area of Patagonia in the transition between Patagonia and Monte phytogeographical provinces. In addition, a disjunct distribution of a haplotype could be detected in three species in south-western Santa Cruz province (Fig. 4B, black dots).
Areas where species may have persisted in situ have been proposed for eleven major regions: (I) Chilean-Coastal range north of 42°S and Chiloé island, (II) a longitudinal zone located in the eastern flanks of the Andes from 38°S to 42°S, (III) in scattered areas located in the high-Andean region north of 38°S, (IV) along the Atlantic coast of Rio Negro Province (approximately 41°S), (V) San Jorge Gulf (approximately 46°S to 47°S), (VI) Southern Santa Cruz (approximately 50°S to 52°S); and (VII) Tierra del Fuego (approximately 53°S); and also in inland Patagonia: (VIII) central Chubut Province, (IX) central northern Santa Cruz Province, (X) south-eastern Santa Cruz Province, and (XI) Tierra del Fuego (Fig. 4B). Secondary contact zones have been identified (a & b) on the eastern flanks of the Andes (c) in Central Patagonia (approximately 46°S) and (d) along a diagonal following the course of the Río Chico basin in Santa Cruz Province (approximately 48°S to 49°S; Fig. 4B). Recent demographic expansions have been hypothesized in all reviewed studies and multiple directions have been proposed for them: from the east towards the southwest at different latitudes; with a north–south direction from southern Neuquén (approximately 39°S) and northern Chubut (approximately 42°S) along the Andes, and from south-eastern Chubut across the Steppe; additional expansion routes, but with a south–north direction from southern and northern Santa Cruz, southern Chubut, and southern Rio Negro Province, have been also proposed (Fig. 4B).
In this review of phylogeographical patterns for Patagonian terrestrial fauna and flora, we describe the location of potential glacial refugia and phylogeographical breaks, and areas of species expansion and colonization routes, some of which are consistent across a number of species; we also distinguish geographical areas within Patagonia in need of further research efforts. Indeed, the majority of focal plant taxa are Andean forest species with economic value (Pastorino & Gallo, 2002; Marchelli & Gallo, 2006; Acosta & Premoli, 2010; Arana et al., 2010), with the exception of two non-woody steppe species (Tremetsberger et al., 2009; Cosacov et al., 2010). Increased research efforts are needed to broaden the taxonomic and geographical coverage of herbaceous Andean species, as well as woody Steppe species. Terrestrial vertebrate taxa are limited to mammals and lizards. Lizard studies have been based on denser sampling schemes in central and northern Patagonia, whereas mammal studies have typically included wider geographical coverage, although these are still scarce in the southernmost areas (Pardiñas et al., 2011). Frogs have only recently begun to be studied in a phylogeographical perspective (Nuñez et al., 2011).
Populations with high genetic diversity suggest that they have either persisted for periods long enough for the maintenance of high genetic variation, or that the inhabited area was recolonized by a number of genetically distinct lineages from multiple refugia (Hewitt, 1996; Petit et al., 2002; Mraz et al., 2007). The presence of private haplotypes is crucial to distinguish between these alternatives (Pinceel et al., 2005; Mraz et al., 2007; Ronikier et al., 2008; Huck et al., 2009) because areas with low or null levels of within population genetic diversity but with private haplotypes suggest the existence of long lasting fragmented refugial areas.
The articles surveyed did not suggest a dramatic shift in species distributions during the LGM; a notable contrast with studies in northern continents (Pinceel et al., 2005; Mraz et al., 2007). Nevertheless, reported genetic patterns strongly support the existence of at least six stable areas where species survived during glaciations (i.e. east of the Patagonian Andes, northern Chiloé Island, northern Chilean Coast, Atlantic Coastline, northern Patagonia, and South of Santa Cruz Province and Tierra del Fuego). These areas show higher genetic diversity than areas where post-glacial colonization occurred, as in some areas of the southern Patagonian Steppe. Although vertebrate information is missing for most of the Santa Cruz Province, lizard data (Breitman et al., 2011) identify high genetic diversity areas also demonstrated by plants.
Although more frequently documented in plants, putative refugial areas are also revealed in vertebrate phylogeographical patterns. Peripheral refugia were reported in 70% of the reviewed species (including at least one species each of woody and herbaceous plant, lizards, and rodents), indicating that, in Patagonia, the survival of species near the glaciers seems to be common (in contrast to the Northern Hemisphere; Hewitt, 2004; Pinceel et al., 2005; Huck et al., 2009). Regarding the location of peripheral refugia, the highest coincidence within and between groups was observed in the eastern foothills of the Andes between 39°S to 43°S, an area where the westerlies and associated precipitation peaks shifted to the north during LGM and back southward during interglacial periods; thus, humidity gradients were particularly produced in southern and northern Patagonia, whereas, at middle latitudes, moisture appeared more stable (Mancini et al., 2005). An alternative non-exclusive explanation is that, at those latitudes, the ice sheet was confined to the peak summits with most of the foothill forests remaining unglaciated, thus allowing persistence of different taxa through later Quaternary climate shifts (Heusser, Heusser & Lowell, 1999).
Lowland refugia along the western Chilean Coast were the second most frequent in both vertebrate and plant studies (Victoriano et al., 2008; Quiroga & Premoli, 2010). Less common but also coincident between taxa were lowland refugia along the Atlantic coastline, which today harbour remnants of a probably more extended easterly refugium that could have persisted on what is now the continental shelf, given that the coastline fluctuated in several longitudinal degrees to the east during glacial cycles (Ponce et al., 2011). An interesting pattern found in all plant studies with sampling that covered the entire latitudinal range of Patagonia are the potential lowland refugia inferred as far as 51°S, both in the Andean and in Steppe areas, suggesting stable refugia at high latitudes (Jakob, Martínez-Meyer & Blattner, 2009; Tremetsberger et al., 2009; Cosacov et al., 2010; Mathiasen & Premoli, 2010). The existence of high genetic diversity, private haplotypes, together with the fact that all instances correspond to species with low dispersal capabilities, strongly suggests this shared pattern is the result of in situ persistence rather than a recolonization from northern refugia. Further supporting this hypothesis, divergent lineages with high genetic diversity within them have been found in lizards at these same latitudes (Breitman et al., 2011). Less common are valley-type refugia north of 36°S, a region characterized by an Alpine valley-type glaciations (Rabassa, 2008b) that could have favoured the persistence of small, isolated populations (Himes et al., 2008; Azpilicueta, Marchelli & Gallo, 2009; Cosacov et al., 2010; Mathiasen & Premoli, 2010).
In summary, three salient common patterns in the geographical distribution of potential refugia could be retrieved from the available data: valley-type refugia north of 36°S, multiple concordance of peripheral refugia along the eastern foothills of the Andes between 39° to 43°S, and lowland refugia as far as 51°S.
Phylogeographical breaks and possible past scenarios
Concordance in the distributions of divergent lineages or phylogroups among several Patagonian taxa appears to exist at a large spatial scale, although their temporal concordance is difficult to assess given that dating is susceptible to many sources of error, particularly if no fossil record is available for calibration. Of the surveyed studies, very few have calculated divergence times with a molecular clock (Cosacov et al., 2010); instead, lineage divergence time was estimated in relation to the dating of major geological and climatological events. Unfortunately, the latter practice ties cause to effect rather than enabling independent dating of divergences, such as from molecular data.
Miocene–Pliocene Andean orogeny and associated tectonic processes that occurred at 35° to 38°S may have had a major role in shaping the evolutionary history of many taxa. Indeed, several plant species or genera appear to have diversified in these northern areas and then migrated south and southeast, colonizing new niches that emerged concomitant with the uplift of the Andes and the consequent desertification of the Steppe (Barreda, Guler & Palazzesi, 2008; Ramos & Ghiglione, 2008). This scenario was proposed for C. polyrhiza (Cosacov et al., 2010), Hordeum (Jakob et al., 2009) and H. incana (Tremetsberger et al., 2009).
Within the younger scenarios temporally situated in the Quaternary Pleistocene glacial cycles, dynamics of the river basins, palaeobasins, and coastline shifts (Hernández, González & Hernández, 2008; Martínez & Kutscher, 2011; Ponce et al., 2011) constitute possible factors that may have moulded the evolutionary history of the Patagonian lineages. Indeed, on the Patagonian Steppe, several breaks have been suggested to be concordant with river basins that currently drain (or that in the past drained) the Andean East watershed by crossing the steppe to the Atlantic Ocean. These include the following rivers: Río Colorado (Liolaemus elongatus; Morando, Avila & Sites, 2003), Río Neuquén (L. elongatus; Morando et al., 2003), Río Limay (Liolaemus bibronii; Morando et al., 2007), Río Chubut (Phyllotis xanthopygus; Kim et al., 1998; Hordeum; Jakob et al., 2009), Río Coig (Hordeum; Jakob et al., 2009) and Río Chico (C. polyrhiza; Cosacov et al., 2010). A rarely considered process in the Patagonian Steppe also could have molded biodiversity patterns such as Quaternary river basins that converted Patagonia into a highly fragmented landscape, as revealed today by extensive glaciofluvial gravel accumulations throughout the Steppe region revealing ancient powerful currents discharging to the sea (Martínez & Kutscher, 2011) (Fig. 4, inset).
Coastline shifts along the Atlantic could have promoted a vicariant pattern shared by three Liolaemus species at the northern and southern rim of the Somuncurá Plateau (Morando et al., 2003, 2007; Avila et al., 2006). When the coastline moved west, some populations migrated to the north, whereas others to the south of the Plateau promoting vicariance and reciprocal monophyletic haplogroups in these taxa.
In summary, major genetic discontinuities in terrestrial Patagonian organisms follow a clear latitudinal pattern. These breaks, located along the Andes and on the steppe, appear to be the result of processes operating at different temporal scales; the Andean breaks appear to be more ancient and likely associated with Pre-Quaternary processes (Andean orogeny and palaeobasins), whereas the more recent stepparian breaks are probably the result of Quaternary events such as glaciations cycles, as well as the associated variations in the dynamics of river basins and coastline shifts.
A target of future studies should be the inclusion of molecular clock calibrations based on fossils to elucidate ages of species divergent events, which may aid to discern if recent phylogeographical breaks only reflect geographical structuring with no underlying historical barriers.
Past demography: range expansion and colonization routes
Some northward expansions also occurred from central Santa Cruz to southern Chubut (Jakob et al., 2009; Cosacov et al., 2010) and from Chubut to northern Rio Negro Province (Morando et al., 2004; Jakob et al., 2009). Several instances of allopatric isolation followed by range expansion may have produced some of the areas of secondary contact we observe today, which are most probably of recent origin, after the GPG or LGM (Palma et al., 2005; Azpilicueta et al., 2009; Cosacov et al., 2010; Lessa et al., 2010). These secondary contact areas suggest that some species survived in more than one refugium, as previously proposed for the eastern Andes by Premoli (1998). Contact areas appear thus far not to be congruent between animals and plants. Variation in dispersal and differentiation rates among species may be responsible for the different location of most of these areas.
Although much work remains to be done, concordant dispersal routes support the expected outcome of niche tracking following ice sheet fluctuations and climate change. Other concordant patterns suggest south–north expansions and in situ persistence in the southernmost areas. This may indicate that Pleistocene climate change and ice sheet fluctuations were not always the most direct factors acting on the demographic dynamics of Patagonian organisms, rather they acted in concert with other factors in more complex scenarios. It is noteworthy that several organisms presumably have remained isolated in patchy populations in the northern Andean area. Several non-mutually exclusive factors may explain this pattern such as the more intricate topography of the Andes favouring isolation among valleys and slopes, the narrow forested area which is more easily split into small areas, and the characteristic valley glaciation (Rabassa & Clapperton, 1990) that could have disrupted gene flow across valleys favouring local population persistence. Although there is lower incidence of fragmentation on the steppe, some aspects of its geological history (e.g. Quaternary palaeobasins, coastline shifts) allow us to expect that fragmentation might have taken place more often than has yet been detected.
This contribution is the result of the symposium ‘Palaeogeography and palaeoclimatotology of Patagonia: Effects on Biodiversity’ held at La Plata Museum in May 2009. The authors are indebted to Enrique Lessa, one other anonymous referee, and the editors Daniel Ruzzante and Jorge Rabassa for their insightful comments on the manuscript. We also thank other members of the Grupo de Herpetología Patagónica, the BYU Kennedy Center for International Studies, Department of Biology, and the Bean Life Science Museum, as well as Dr Xavier Scheldeman and Maarten Van Zonneveld for valuable tips about GIS analyses. Financial support was provided by the grants: NSF-PIRE award (OISE 0530267) for support of collaborative research on Patagonian Biodiversity granted to the following institutions (listed alphabetically): Brigham Young University, Centro Nacional Patagónico (A.R.), Dalhousie University, Instituto Botánico Darwinion (A.R.), Universidad Austral de Chile, Universidad de Concepción, Universidad Nacional del Comahue, Universidad Nacional de Córdoba, and University of Nebraska. PICT 2006-506 ANPCYT-FONCYT (L.J.A.), ANPCYT-FONCYT 33789 (M.M.), PIP 2009-12 (M.M.). CONICET PIP 11220080101264 (A.S.N. & A.A.C.), FONCyT PICT 01-10952 and 01-33755 (A.S.N. & A.A.C.), and SeCyT-UNC 05/I625 and 05/I648 (A.S.N. & A.A.C.).
Table A1. Lists of the 33 articles included in the present study
Number of individuals
Number of sites
Diversity index (DI) obtained from the respective article is indicated. Ar, allelic richness; Sh, Shannon index; HE, expected heterocigocity; HD, haplotype diversity. D, available in the revised article, C, calculated for the present study.