While the various Nothofagus species and their putative hybrids differ substantially in morphology, there were no diagnostic genetic differences within our samples across the three chloroplast genes and eight nuclear microsatellite DNA markers. In particular, N. pumilio and N. antarctica possess considerable overlap in cpDNA and nuclear microsatellite population composition. The other three species of Nothofagus (N. betuloides, N. dombeyi and N. nitida) also had a substantial overlap in population structure.

Nothofagus pumilio and N. antarctica have distinct phenotypes

Nothofagus pumilio and N. antarctica have distinct leaf phenotypes that match the descriptions in the existing literature [16]. While they are similar in leaf area, phenotypically pure N. pumilio leaves have crenated edges with round lobes that are separated by numerous veins, while N. antarctica leaves have serrated edges and fewer veins (Fig. 2B and Fig. 4). Such phenotypes explain why the N. pumilio leaves have a significantly lower ridge-to-vein ratio and ridge-to-area ratio than N. antarctica leaves (Fig. 5).

Putative Nothofagus hybrids have intermediate leaf phenotypes

We identified both parental species and their putative hybrids by morphology. We thus start with a caveat that while we did not use fluctuating asymmetry to assign individuals as putative hybrids (we focused on selecting trees with leaves that had a distinct morphology from N. pumilio and N. antarctica), fluctuating asymmetry could have influenced our decisions. If the species are indeed overlapping genetically as microsatellite and chloroplast DNA sequences suggest, their morphological differentiation would be potentially caused by few loci or by phenotypic plasticity.

We found no significant differences in the ridge-to-vein ratios between N. antarctica and the putative hybrid leaves. However, the mean leaf ridge-to-area ratio of the putative hybrids is significantly higher than that of N. pumilio and significantly lower than that of N. antarctica (Fig. 5). Putative Nothofagus hybrids therefore have an intermediate leaf morphology distinguishable from both parental species. This observation is consistent with previous observations of the N. pumilio and N. antarctica hybrids at 41ºS displaying intermediate numbers of leaf marginal lobes as well as the findings in other plant genera, such as Arabidopsis and Quercus [5, 7, 19, 26, 51]. Our characterization of intermediate leaf morphology adds data to the morphological classification of Nothofagus hybrids, complementing previous findings that demonstrated intermediate characteristics in bark roughness and color, stem straightness, and crown form in putative hybrids of N. pumilio and N. antarctica [47].

Putative Nothofagus hybrids have higher leaf fluctuating asymmetry

Although the leaf morphology of putative hybrids was intermediate between the two parental species, their leaf fluctuating asymmetry was significantly higher than that of the parental species, N. pumilio and N. antarctica (Fig. 6). Such high fluctuating asymmetry was present for both the leaf length and leaf width. The higher fluctuating asymmetry of hybrid individuals has been attributed to developmental instability caused by the disruption of coadapted gene complexes during hybridization [8, 23]. Indeed, our linear regression between the fluctuating asymmetry values of leaf length and leaf width showed a positive correlation for only the putative hybrids, indicating that the factors that drive asymmetry operate on width and length simultaneously (Fig. 7) and suggesting that the two could be caused by underlying developmental disruptions in hybrids.

We also showed that the leaf area of hybrids had a negative correlation with the fluctuating asymmetry of leaf length and width: larger leaves were less likely to be asymmetrical (Fig. 8). In contrast, the N. pumilio leaf area was negatively correlated only with the fluctuating asymmetry of leaf width, while the N. antarctica leaf area had a positive correlation only with the leaf length fluctuating asymmetry. In other words, larger N. pumilio leaves are likely to be less asymmetrical in width, while larger N. antarctica leaves are likely to be more asymmetrical in length. However, since the adjusted R2 values for all of the models were very low (the highest R2 value being 0.2009), there may be external factors that significantly affect both leaf size and the fluctuating asymmetry of leaf length and width for all of the study species and their hybrids.

Nothofagus pumilio and N. antarctica have identical chloroplast DNA

While cpDNA sequences in GenBank show differences between N. pumilio and N. antarctica species, our N. pumilio and N. antarctica samples (populations S1-S3, 52°S) possessed identical cpDNA sequences for all genes (Fig. 9). Furthermore, our samples differed from the conspecific GenBank cpDNA sequences (see results for details). A possible explanation for the identical chloroplast sequences between phenotypically distinct N. antarctica and N. pumilio individuals in the 52°S region is chloroplast capture through introgression. Indeed, introgression events have been documented in Argentine and Chilean Nothofagus populations near our collection sites [1].

Studies have hypothesized that backcrosses to N. pumilio or N. antarctica might have caused such introgression [1, 40, 48]. Our putative hybrids (if correctly identified) resembled N. antarctica morphologically, perhaps suggesting backcrosses towards N. antarctica (no significant difference in leaf ridge-to-vein ratio between N. antarctica and putative hybrids,Fig. 5). This supports the findings of Quiroga et al. [42], where no difference was found for maximum width, total and minimum length, and the ratio of total and minimum length between N. antarctica and hybrid leaves. Similar introgression patterns have also been observed between N. nervosa and N. obliqua, which exhibit asymmetric late-generation backcrosses towards N. obliqua at high altitudes and backcrosses in both directions at lower altitudes [13]. El Mujtar et al. 13 suggest that the preference towards N. obliqua backcrossing at high altitudes might be due to a selective advantage over other hybrids. Considering that N. antarctica is a generalist species with high adaptive capacity, while N. pumilio prefers well-drained soil, the backcrossing at 52°S might have given hybrids an adaptive advantage, perhaps during unfavorable or variable climatic periods such as continental glaciations [40, 47].

However, these results should be taken with caution, as it has been proven that the use of chloroplast DNA to study phylogenetic relationships between plants can lead to inaccurate conclusions–nuclear DNA analysis as well as ample taxon sampling is needed to properly elucidate the relationships between the Nothofagus species [49].

Low nuclear genetic differentiation between species

Previous studies of nuclear ITS regions from sympatric Nothofagus populations identified N. antarctica as a sister clade to the evergreen species N. betuloides, N. dombeyi, and N. nitida, while N. pumilio diverged earlier [1, 39]. These results suggest that the 5 morphologically distinct species also have distinguishable genomes and well-supported phylogenetic relationships based on nuclear gene data. However, our analyses using eight unlinked variable loci revealed unexpectedly low population structure between species and across most populations. In every STRUCTURE analysis, the calculated values of K [14] were lower than the number of morphologically defined species or populations. The best value of K for all populations/species was 4 rather than 5 (the total number of species) (Fig. 10). K was 3 for N. pumilio, N. antarctica, N. betuloides, and N. dombeyi within the southern populations (Fig. 11) and only 1 for the southern N. pumilio and N. antarctica populations (Fig. 12). While the morphological differences in our samples might be due to phenotypic plasticity, we find this unlikely since many of the Nothofagus species are sympatric, and we collected samples from different species within the same tree stands.

While there was variation within and between populations, STRUCTURE showed that every species had individuals assigned to populations in common with other species (Figs. 10, 11 and 12). Moreover, though past studies classify N. pumilio and N. antarctica as clearly distinct species [48], we were unable to tell sympatric (51–53°S) N. pumilio and N. antarctica apart with microsatellite markers (STRUCTURE, Fig. 12, best K = 1) despite high genetic diversity within each species. The FST between N. pumilio and N. antarctica was lower than the FST calculated within populations of any of the Nothofagus species (Table 3). As found by Premoli et al. [40], such results could be explained by the recurrent hybridization and divergence between N. pumilio and N. antarctica through glacial–interglacial cycles. Another possible explanation is that phenotypic differences between these otherwise genetically homogenous populations could be driven by only a small proportion of the Nothofagus genome. For example, this surprising genome homogeneity between morphologically different species has been observed in the seed finch species Sporophila plumbea and S. beltoni, which differ in plumage and beak color but have only six highly differentiated loci [36], suggesting that assortative mating within morphotypes is not strong enough to prevent genome mixing. Our analyses suggest that a similar scenario might be true for N. pumilio and N. antarctica, particularly within the 51–53°S region. Future studies with more extensive genetic sequencing will be needed to clarify the extent of genetic differentiation between these Nothofagus species.

While it is possible that introgression mixed up sympatric populations, it is notable that even distant collection sites (41–42°S, 51–52°S, 53°S) of N. pumilio and N. antarctica had very low differentiation (Table 3), with large overlaps in STRUCTURE population assignment. Although we sampled few individuals from the northern populations (1 N. pumilio and 2 N. antarctica from 41–41°S), this result was unexpected because prior studies showed clear population differentiation between the northern and southern Nothofagus populations [33, 46, 48]. Therefore, sequencing more northern N. pumilio and N. antarctica individuals would improve the accuracy of our analysis and confirm whether the lack of genetic differentiation between the two species is limited to the more southern regions near 51–53°S.

While we did not quantify morphological differences for the three other Nothofagus species used in our study–the evergreen N. betuloides, N. dombeyi, and N. nitida–their leaves are easily identifiable. Nothofagus betuloides has leaves much thicker and smaller than those of N. pumilio and N. antarctica, and both N. dombeyi and N. nitida have thick yet bigger and more pointed leaves than N. betuloides. Nothofagus nitida leaves were collected exclusively on Chiloé Island (population N3) and were larger than N. dombeyi leaves. However, we cannot exclude the possibility that some N. dombeyi individuals were mistaken for N. nitida since the two species were less morphologically distinct, and STRUCTURE analyses overlap in population assignment for N. dombeyi and N. nitida (Figs. 2B and Fig. 10). Previous studies have also reported evidence of hybridization between the deciduous N. antarctica and the evergreen N. dombeyi at 42°S, which could explain the genetic similarities observed between N. antarctica and N. dombeyi in our analysis at 51–53°S [50]. If the similarity between species was indeed caused by introgression, then our data also supports evidence of a previously undocumented hybridization between N. antarctica and another evergreen species, N. betuloides, explaining the substantial genetic overlap between N. antarctica, N. betuloides, and N. dombeyi (Fig. 11).

Hybridization between tree species may serve as an evolutionary strategy for successful recolonization following habitat disruptions. In South America, Nothofagus forests have experienced repeated incidents of environmental change–such as marine transgressions and glaciations–during which hybrid individuals may have possessed advantageous traits that enhanced their survival. Today, similar large-scale disturbances continue to affect Chilean forests, particularly through recurrent wildfires since the Last Glacial Maximum. In this context, assessing the extent of Nothofagus hybridization and understanding how species admixture affects an individual’s developmental instability or the genetic health of the population may offer insights into the adaptive capacity and resilience of these species under changing environmental conditions.