The work, published in Frontiers of Biogeography, focuses on Antrodiaetus, a genus of fossorial spiders with no ballooning dispersal, lifespans estimated at 15 to 30 years, strong habitat preferences, and a nearly nonexistent fossil record. The only fossil representing the family, Cretacattyma raveni, was found in Mongolia and dates to the Cretaceous Period, anchoring the group’s minimum age but offering little help for dating splits among living species. Because burrowing spiders rarely become trapped in sediment layers, the researchers turned to the landscape itself as a clock.

The centerpiece of the study is the Pacificus species complex, a group of cryptic spider lineages distributed disjunctly across the Pacific Northwest. One set of lineages inhabits the Coastal and Cascade mountain ranges from northern California to Alaska, while others live in the interior forests of the Rocky Mountains in Idaho. Between them lies the inhospitable Columbia Basin and Oregon Highlands, a dry expanse that separates mesic forest communities on either side. Decades of research on other dispersal-limited plants and animals across this same break have produced two competing explanations: an Ancient Vicariance hypothesis, in which mountain uplift and subsequent drying fragmented a once-contiguous forest, and an Inland Dispersal hypothesis, in which spiders moved eastward through temporary mesic corridors opened at the edges of retreating Pleistocene glaciers.

Critically, the two hypotheses make different predictions about the shape of the evolutionary tree. Ancient Vicariance implies that coastal populations should be more closely related to each other, with inland populations representing older, isolated refugia. Inland Dispersal implies that inland populations should nest within or sister to particular northern or southern coastal populations, depending on the route of dispersal. Because each hypothesis carries its own timeframe—millions of years for the ancient scenario versus thousands for the recent one—choosing between them also provides a date that can calibrate a molecular clock.

To generate the genomic data needed to test these ideas, the team employed ultraconserved elements, or UCEs—thousands of loci captured using a spider-specific probe kit and sequenced with Illumina technology. UCEs have proven highly informative in mygalomorph systematics, but not all loci behave equally well for dating analyses. The researchers applied a “gene shopping” strategy, using the SortaDate software to rank 1,671 recovered loci by root-to-tip variance, agreement with the preferred tree topology, and overall tree length. Starting from a 95 percent occupancy matrix of 71 loci, they distilled the dataset down to the top 20 most clock-like loci, trading breadth of genomic coverage for a cohesive, consistent evolutionary signal.

Phylogenetic analyses using both concatenated supermatrices in IQTree2 and coalescent-based species trees in ASTRAL-III confirmed the major species groupings within Antrodiaetus. A phylogenetic network built with SplitsTree revealed significant discordance among loci at certain deep nodes, particularly within the Inland Pacific Northwest group, likely the result of rapid ancient radiations that left too little sequence divergence for even genomic-scale data to fully resolve. Despite this noise, one pattern held across every analysis: the enigmatic “pacificus SE” lineage, a small population in the southern Rocky Mountains, was consistently sister to the northern and southern “pacificus WEST” populations—the exact topology predicted by Ancient Vicariance.

Hypothesis testing then became a formal statistical exercise. The researchers compared eight models in BEAST 2’s StarBeast3 package, crossing strict and relaxed molecular clocks with Yule and Fossilized Birth-Death branching models and with either the ancient or recent calibration. Nested sampling ranked a relaxed-clock, Fossilized Birth-Death model with the ancient calibration as the best fit, but relaxed-clock runs failed to reach stationarity or convergence, producing trees with negative branch lengths and implausible topologies. Bayes factor comparisons told a clearer story: the Ancient Vicariance model was strongly supported over Inland Dispersal under both clock models, with Bayes factors of 25.7 under a strict clock and 10.5 under a relaxed clock. The team therefore adopted the next-best model—a strict clock with the ancient calibration and a Fossilized Birth-Death branching process—which converged reliably.

Under that model, the researchers estimated a mean strict-clock rate of 4.499 × 10⁻⁴ substitutions per site per million years for Antrodiaetus and its sister genus Atypoides, with a 95 percent highest posterior density interval spanning roughly 2.7 to 6.3 × 10⁻⁴. This rate is an order of magnitude slower than nuclear rates previously applied to spiders, a difference the authors attribute largely to life history. Mygalomorphs mature slowly and cannot balloon, limiting gene flow and generation turnover compared with araneomorph spiders, whose faster metabolisms and dispersal produce quicker molecular evolution. The conserved nature of the UCE core regions and the deliberate selection of clock-like loci likely also contributed to the slower estimate.

The resulting chronogram reshapes understanding of Pacific Northwest biogeography. The split between Antrodiaetus and Atypoides was placed around 138 million years ago, in the Late Jurassic or Early Cretaceous, when a warm, ice-free Holarctic likely offered continuous habitat. Within the genus, major divergences occurred during the Eocene as Rocky Mountain uplift and global cooling increased seasonality across western North America. Most strikingly, the separation of “pacificus SE” from “pacificus WEST” was dated to approximately 8.5 million years ago—older than the Pliocene timeframe of the classic Ancient Vicariance hypothesis but consistent with the Cascade orogeny, suggesting that early uplift initiated the isolation before regional drying sealed it. This makes the spider split the oldest yet documented across the Columbia Basin biogeographic break.

The study also links the genus’s diversification to the rise of North American grasslands. The disjunct eastern distribution of A. robustus, embedded within a western species group, corresponds with the Miocene retreat of transcontinental forests and their replacement by prairie. Similarly, the divergence of A. apachecus in the American Southwest appears tied to the isolation of montane “sky island” habitats as the region dried. The Japanese species A. yesoensis, deeply nested among North American lineages, points to dispersal across warm, forested northern land bridges during the Paleogene before climate change severed the connection. Together, these findings demonstrate that biogeographic events can serve as rigorous, testable calibrations when fossils fail, and they provide the first UCE-based evolutionary rate for atypoid spiders—a tool the authors hope will illuminate divergence times across the entire mygalomorph radiation.

Subject of Research: Using biogeographic calibrations to estimate divergence times and evolutionary rates in the folding-door spider genus Antrodiaetus where fossils are scarce

Article Title: Hypothesis-driven biogeography to estimate rates and dates when fossil evidence is scarce: a case study using an ancient, Holarctic genus of spider ( Araneae , Antrodiaetidae , Antrodiaetus )

Article References: Ciaccio, E., & Hamilton, C. A. (2026). Hypothesis-driven biogeography to estimate rates and dates when fossil evidence is scarce: a case study using an ancient, Holarctic genus of spider (Araneae, Antrodiaetidae, Antrodiaetus). Frontiers of Biogeography, 19, Article e182722. https://doi.org/10.21425/fob.19.182722

Image Credits: AI Generated

DOI: 10.21425/fob.19.182722

Keywords: Antrodiaetus, biogeography, molecular clock, ultraconserved elements, mygalomorph spiders, ancient vicariance, Pacific Northwest, divergence dating, phylogenomics, Columbia Basin, fossil record, BEAST 2

The research team collected scale-leaf tissue from M. callistoma in Shiyan, Hubei Province, and sequenced it on an Illumina HiSeq 4000 platform at a mean depth of over 900-fold. Assembly with GetOrganelle and polishing with Pilon produced a single circular plastome of 29,784 base pairs with a GC content of 30.8 percent. That figure might sound unremarkable until compared with a typical photosynthetic plant, whose plastome runs roughly 150,000 base pairs and carries more than a hundred genes. The M. callistoma plastome contains just 40 genes: 20 protein-coding genes, 16 transfer RNAs, and four ribosomal RNAs. Strikingly, the canonical quadripartite structure found in almost all land plant plastids, with two inverted repeats separating large and small single-copy regions, is entirely absent. No inverted-repeat boundaries could be detected at all.

The gene inventory tells the story of a genome in terminal decline. Every photosynthetic function has been erased. There are no photosystem subunits, no NADH dehydrogenase genes, no cytochrome b6f complex components, no ATP synthase subunits, and no rubisco large subunit. What remains is essentially a minimal housekeeping kit: ten genes for small ribosomal subunit proteins, seven for large subunit proteins, and just three other coding genes, infA, matK, and accD. Four of the retained genes carry a single intron. This is the genomic signature of an organelle whose core reason for existence, photosynthesis, has been outsourced to fungal partners, leaving relaxed selection to strip away everything the plant no longer needs.

To place this genome in context, the researchers compiled published plastomes from 30 mycoheterotrophic species across six families, ranging from orchids such as Gastrodia and Lecanorchis to the bizarre holoparasites Sciaphila and Mitrastemon. The comparison revealed a spectrum of decay. Sciaphila thaidanica and Mitrastemon yamamotoi carry plastomes of only about 20 kilobases, among the smallest known in flowering plants, while at the other extreme some Burmannia species retain genomes approaching 100 kilobases. GC content showed a weak but statistically significant positive correlation with genome size, suggesting that as these genomes shrink, their nucleotide composition shifts in parallel. Within Monotropoideae itself, Pterospora andromedea holds the largest plastome at roughly 40 kilobases, while the more recently diverged Monotropa uniflora has the smallest at about 27 kilobases.

Whole-genome alignments exposed another dimension of decay: the collapse of synteny, the conserved order of genes along the chromosome. Among Monotropoideae species, gene order is scrambled, particularly in Hypopitys, Monotropsis, Monotropastrum, and Monotropa. Fragment loss has accumulated in more recently diverged lineages, so that even within the genus Monotropa, where synteny is relatively well preserved, the overall picture across the subfamily is one of progressive disintegration. Ancestral-state reconstruction across the 30 species showed that losses of the large genes ycf1 and ycf2 occurred convergently among eudicot lineages, while all sampled Monotropa species lacked rps23, pointing to a lineage-specific loss within that genus.

The phylogenomic analysis went beyond genome description to resolve long-standing questions about relationships within Ericaceae. Using a concatenated alignment of plastid protein-coding genes nearly 48,000 base pairs long, with Actinidia chinensis as an outgroup, the team recovered a well-supported split into two major clades. Clade 1 united Pyroloideae, Arbutoideae, and Monotropoideae, with Pyroloideae branching earliest and Monotropoideae emerging as the youngest lineage. Within Monotropoideae, the sampling covered 10 of the 12 recognized genera and achieved maximum bootstrap support of 100 for the backbone. Pterospora and Sarcodes occupied successive early-branching positions, followed by Pleuricospora, while a clade of Monotropsis, Monotropastrum, and Monotropa was sister to a group containing Allotropa, Hemitomes, Pityopus, and Hypopitys. Clade 2 comprised the core Ericaceae, subdivided into Vaccinioideae and Ericoideae.

Divergence dating with fossil calibrations placed the split between Pyroloideae and the remaining subfamilies at approximately 67.77 million years ago, in the latest Cretaceous. Arbutoideae and Monotropoideae diverged around 66.57 million years ago, and diversification within Monotropoideae began about 63.35 million years ago, immediately after the end-Cretaceous extinction. The genus-level splits came much later: Sarcodes and Pterospora separated roughly 45.79 million years ago, Monotropsis split from the lineage leading to Monotropa and Monotropastrum about 39.73 million years ago, and those two genera diverged about 26.74 million years ago. The rapid radiation of Rhododendron and Vaccinium, by contrast, occurred only around 1 to 2 million years ago.

Perhaps the most striking finding concerns geography. Ancestral-range estimation using the DIVALIKE+J model in RASP placed the most recent common ancestor of Clade 1 in Asia during the late Cretaceous to early Cenozoic. From there, the lineage dispersed to North America, most plausibly across the Bering Land Bridge, which connected the continents from the latest Cretaceous into the early Paleocene, when a warm, humid belt of deciduous broadleaf forest stretched across the region. Then came a remarkable reversal: the ancestor of the Monotropsis, Monotropastrum, and Monotropa lineage appears to have dispersed back to Asia during the late Eocene to early Oligocene, around 39.70 million years ago, giving rise to the East Asian endemics Monotropastrum humile and Monotropa callistoma. The ghost plants, in other words, crossed the world twice.

The comparative analysis also yielded a testable hypothesis about the order in which plastid genes are lost as plants abandon photosynthesis. Across the 30 species examined, the ndh genes were absent everywhere, followed by photosystem, cytochrome, rubisco, and ycf3/4 genes, then the RNA polymerase genes, then atp genes, and finally the ribosomal, transfer RNA, and housekeeping genes such as accD, clpP, and matK. The sole exception was Burmannia cryptopetala, which retains atp genes, a pattern associated with mixotrophic or partially parasitic lifestyles. The team proposes that the degree of a plant’s absolute dependence on fungal hosts correlates with the breadth of plastid gene loss, with Monotropa, whose plastid genes evolve roughly 13 times faster than its nuclear genes, representing an advanced stage of degradation in which even clpP has been discarded.

Yet the decay is not random, and that is the deeper lesson of the study. Housekeeping genes persist with remarkable consistency: accD, which catalyzes the first committed step of fatty acid biosynthesis and is essential for building the plastid’s own membranes, was retained in every single mycoheterotrophic species examined. The retention of matK across all Monotropoideae and Burmanniaceae, despite its patchy record in orchids, and the lineage-specific survival of ycf1 and ycf2 in some orchids and Burmanniaceae but not others, show that different lineages face different residual functional constraints. The endpoints of plastid genome reduction form a spectrum rather than a single destination, ranging from the streamlined but intact genomes of Monotropoideae to the complete loss of the plastid genome in Balanophora, which has transferred even accD to the nucleus. When an organelle’s core function is outsourced, its genetic system converges, again and again across independent lineages, toward the same extreme simplification. The ghost plants of Asia and North America, silent and white beneath the forest canopy, are carrying out that experiment in real time, and their genomes are the record.

Subject of Research: Plastome evolution and phylogenomics of the mycoheterotrophic subfamily Monotropoideae (Ericaceae)

Article Title: Phylogenomics Clarifies Plastome Reduction and Phylogenetic Relationships in Monotropoideae

Article References: Deng, G., Duan, K., Xiang, N., Gao, F., Zhan, P., Shu, R., Lin, Q., Zuo, Q., & Yuan, T. (2026). Phylogenomics Clarifies Plastome Reduction and Phylogenetic Relationships in Monotropoideae. Ecology and Evolution, 16(10), Article e74449. https://doi.org/10.1002/ece3.74449

Image Credits: AI Generated

DOI: 10.1002/ece3.74449

Keywords: Monotropoideae, mycoheterotrophy, plastome reduction, Ericaceae, Monotropa callistoma, phylogenomics, Bering Land Bridge, biogeography, gene loss, relaxed selection, ghost plants, divergence dating