Patagonia in the genome. How Austrocactus relationships were read

Plain, rock crevice and collection

On the sandy and gravelly plains of Santa Cruz, Austrocactus aonikenkensis ranges in habit from low and prostrate to erect. In rock crevices east of Domuyo, A. pauxillus displays almost globose stems up to 3 cm high above ground. In a collection, the same species can grow much larger. Some traits on which initial identification is based therefore change between the field and a pot. The flower, fruit, spine structure and documented origin add information less dependent on stem size alone. [Böhnert et al. 2025]

Böhnert and colleagues combined this assessment with DNA analysis. The revision covers herbarium specimens, name-bearing type material, long-term field observations and plants from collections. After identifications were organised, 764 occurrence records were compiled. Many observations lacked a separate documenting herbarium specimen, so the map and genomic samples represent two complementary scopes of data. The map shows the extent of fieldwork, while sequences allow selected, precisely identified plants to be compared. [Böhnert et al. 2025]

Genome fragments and two ways to construct a tree

The analysis included 44 Austrocactus samples representing species and morphological forms. Material came mainly from first-generation plants grown from seeds collected in Argentina and Chile. Samples were taken from roots or the outer layer of stems. Representatives of Eulychnia and a more distant outgroup of Corryocactus were included to place the relationships within a broader comparison. A single specimen of some species and more numerous samples of widely distributed forms provide uneven resolution at population level. [Böhnert et al. 2025]

The ddRADseq method uses two enzymes that cut DNA at specific sites. Selected fragments scattered across the genome are sequenced. Differences in single nucleotides, called SNPs, serve as comparative markers. The study does not read the entire genome of each specimen. The authors tested different fragment-assembly settings and permitted proportions of missing data. The dataset used to display the main trees contained 84 171 SNPs. A large number of markers helps identify a consistent signal, while stability across settings establishes the reliability of the result. [Böhnert et al. 2025]

The first tree was calculated using maximum likelihood, combining sequence information. The second arose from quartet comparisons in a coalescent model. This model accounts for gene copies having their own divergence histories, which need not all indicate the same species tree. Agreement between the approaches strengthened recognition of the main lineages. Areas of disagreement had biological significance and remained part of the result. [Böhnert et al. 2025]

A lineage with smooth ribs and tuberous roots

Austrocactus spiniflorus from central Chile proved to be the sister lineage to all other species in the genus. It has a robust tuberous root, rhizomes and underground shoots. Ribs on the slender above-ground stems remain smooth, and flowers also arise from older lower areoles. Its mature fruit is yellow and mucilaginous. The other species generally have tuberculate ribs, flowers closer to the apex and fruits ranging from green to dark purple. [Böhnert et al. 2025]

This combination of structure and tree position justified the separate subgenus Spiniflori. The former names Opuntia spiniflora, Erdisia spiniflora and Corryocactus spiniflorus recall earlier difficulties in classifying the plant. Genomic markers confirm its inclusion in Austrocactus. A longer branch on the tree indicates a greater accumulated sequence difference. The figure in the 2025 revision alone does not provide a new molecular-clock estimate of this lineage's age. [Böhnert et al. 2025]

Three lineages in the core of the genus

The remaining species form three main lineages, designated sections in the revision. Coxiani includes A. coxii with the subspecies coxii and colloncurensis, A. nobilis and the newly described A. occidentalis. Their fruits are densely woolly. Section Hiberni includes A. hibernus, A. decorus, A. elegans, A. subandinus, A. pauxillus and plants described as A. longicarpus (A. coxii subsp. longicarpus in POWO). Stigmas in this group are generally yellowish or pale green, and fruits have less wool. [Böhnert et al. 2025] [2-WWW]

Section Austrocactus contains A. aonikenkensis, A. bertinii, A. implexus and four local Andean species, A. diamantinus, A. ferrarii, A. philippii and A. praecox. Its plants are distinguished by deep-purple stigmas and fruits with less wool. The agreement of some morphological traits with DNA-defined groups gives these traits diagnostic value. Stigma colour does not, however, establish a proven mode of pollination. The authors note that the genus's reproductive biology remains poorly understood. [Böhnert et al. 2025]

The first two lineages have smaller ranges near the Andes, with Coxiani occurring farther south than Hiberni. Section Austrocactus has a broader range, extending from Mendoza to southern Patagonia. Partial range overlap creates space for both geographic isolation and contact between plants belonging to different lineages. The revision establishes this spatial pattern, leaving the detailed causes of range formation to be tested using biogeographical models. [Böhnert et al. 2025]

Short branches and gene history

Within the core of Austrocactus, tree branches are short, corresponding to small differences between recently diverged lineages. The authors interpret this pattern as relatively rapid diversification. An additional principal component analysis showed specimen similarity based on independently selected markers. In several taxa, including A. nobilis, A. pauxillus and plants in the longicarpus group, the similarity pattern was more complex than a single division into three sections. [Böhnert et al. 2025]

The discrepancy may result from incomplete lineage sorting. In this process, DNA variants present in ancestors persist after populations separate and are arranged differently in successive genes. Another explanation is gene flow through hybridisation. The revision considers both possibilities but does not distinguish them through a separate test of admixture history. The tree and similarity analysis indicate where to seek further data, maintaining a clear boundary between the observed result and the proposed mechanism. [Böhnert et al. 2025]

A species in a contact zone

The newly described Austrocactus implexus grows in rocky areas of southern Río Negro and northern Chubut, in the zone between the main ranges of A. bertinii and A. aonikenkensis. In the maximum-likelihood analysis, it was closer to the first species, and in the coalescent analysis to the second. Long pale spines and other traits allow it to be recognised in the field. Seed-grown offspring retained this combination of traits. [Böhnert et al. 2025]

The authors proposed a stabilised hybrid origin. The basis is the combination of different positions on the trees, a zone of geographic contact and stable morphology. This result supports the hypothesis, while proving the precise contributions of both parental lineages requires denser population sampling and studies of gene flow. The species remains accepted in both the revision and POWO, with clearly defined uncertainty about its origin. [Böhnert et al. 2025] [2-WWW]

Seventeen species in the revision, sixteen in the database

Five new species were described in the revision, A. decorus, A. diamantinus, A. elegans, A. implexus and A. occidentalis. Current POWO accepts all five names. The difference in species counts concerns longicarpus. In POWO it is a subspecies of A. coxii, while the revision's authors recognise a separate species belonging to a different main lineage from the nominate subspecies of A. coxii. The atlas profile presents this difference under the relevant subspecies, retaining the genomic analysis results. [Böhnert et al. 2025] [2-WWW] [2-WWW]

Distinguishing the database's position from that of a particular revision preserves the biological meaning of the data. An experiment or description labelled A. longicarpus concerns plants from that group. It does not become a description of all three subspecies of A. coxii. Similarly, the former name A. gracilis requires checking whether it denotes actual material from the Nahuel Huapi area or was misapplied to pink-flowered plants from Zapala. The contemporary name organises the description, while historical identification helps reconstruct its scope. [Böhnert et al. 2025] [2-WWW] [2-WWW]

Dry landscapes and the rate of diversification

The sister genus Eulychnia consists mainly of large cacti from the deserts of western South America, while Austrocactus is associated with Patagonia and the neighbouring Andes. In Merklinger and colleagues' 2021 molecular-clock analysis, the common ancestor of the two genera was estimated at about 6,70 million years ago, with a broad 95% HPD interval of 3,28–10,51 million years. HPD describes the age range with the highest probability density in the model used. This dating used plastid sequences and secondary calibration of the age of the Cactaceae family. [Merklinger et al. 2021]

The 2025 revision confirms the sister relationship but does not construct a new full model of dating and historical ranges for Austrocactus. The authors compare the late diversification of the genus's core with the earlier development of dry environments in Patagonia. Temporal succession alone does not identify the event that separated particular populations. Hypotheses concerning Andean uplift, the emergence of a forest belt in central Chile and climate change require comparison of genomes with past environmental conditions. Current data indicate both lineages with a longer independent history and recent splits in which traces of shared ancestors and possible hybridisation remain readable. [Böhnert et al. 2025]