Forgotten borage species returns from synonymy as DNA and nutlets seal the case

Deep in the clay steppes surrounding Lake Balkhash in Kazakhstan, a small blue-flowered herb has been hiding in plain sight — not in the field, but in the pages of botanical synonymy. For decades, botanists folded Lappula betpakdalensis into a broader species, Lappula lasiocarpa, assuming that their strikingly similar winged fruits meant they were one and the same plant. A new study, published in the journal PhytoKeys, has now dismantled that assumption with a one-two punch of phylogenomics and classical morphology, formally reinstating L. betpakdalensis as a distinct species. The case offers a vivid reminder that in plant taxonomy, the devil — and sometimes the species — is in the details of the fruit.

The genus Lappula, a member of the borage family (Boraginaceae), comprises roughly 50 to 80 species scattered across Eurasia, North Africa, the Americas and Australia, with its greatest diversity concentrated in Central Asia. Its fruits, called nutlets, are its signature: tiny seeds bearing either marginal barbed glochids — the hooks that let the plant hitchhike on animal fur — or flattened wings. Nutlet morphology has long been the backbone of classification within the genus, which is precisely why the new findings matter. When two species differ in the very character that defines their sections and species, taxonomists sit up and take notice.

The story begins with a taxonomic quirk. In 1953, the Russian botanist Mikhail Popov established the genus Lepechiniella, separating it from Lappula mainly by a shorter gynobase — the central stalk on which the nutlets sit — and by winged nutlets. Later, in 1986, Nabiev rejected this split and transferred all Lepechiniella species into Lappula. Because one name, Lappula balchaschensis, had already been validly published, Nabiev had to coin a replacement name for the transferred species: Lappula betpakdalensis. The name itself nods to the Betpak-Dala desert region of Kazakhstan, though the species is currently known only from the Lake Balkhash area.

Fast forward to 1995, when Zhu and colleagues revised the Chinese members of Lappula. Following Nabiev’s broad circumscription, they accepted the Chinese species Lepechiniella lasiocarpa into Lappula — but they also swept L. betpakdalensis, L. balchaschensis and a variety of L. omphaloides into synonymy under L. lasiocarpa. On the surface, the move seemed reasonable. Both plants produce heteromorphic nutlets — two distinct fruit types, broadly winged and narrowly winged — and both share annual life form, spreading hairs, narrowly ovate calyx lobes shorter than the nutlets, and a style that never overtops the fruit. L. lasiocarpa is so distinctive within the genus, with its saccate nutlet wings curving upward into a pouch-like structure with entire, densely pubescent margins, that it can be identified at a glance.

But a careful re-examination of the type specimens — the reference specimens that anchor each species name — revealed differences that synonymy could not explain. Although both species bear distinctly winged nutlets, the wings of L. betpakdalensis are dentate, their margins toothed, with each tooth typically terminating in a glochid. The wings of L. lasiocarpa, by contrast, are entire and densely pubescent along their margins. In a genus where nutlet wing morphology is a key character for sectional delimitation, this difference clearly exceeds the expected range of intraspecific variation. Corolla size adds a second diagnostic: L. betpakdalensis flowers measure 3 to 4 millimeters across, while those of L. lasiocarpa span 5 to 6 millimeters.

To test whether these morphological differences reflected genuinely separate evolutionary lineages, the researchers turned to phylogenomics. They obtained four samples of L. betpakdalensis from the Moscow University Herbarium (MW) and the Komarov Botanical Institute’s vascular plant herbarium (LE), and sequenced them alongside 64 Lappula samples from an existing phylogenomic framework. Using deep genome skimming — a technique that sequences the whole genome at low coverage and then mines it for organellar and targeted nuclear loci — the team assembled complete chloroplast genomes for all 68 samples and recovered hundreds of Angiosperms353 nuclear loci using the HybPiper pipeline.

The chloroplast genomes, ranging from about 145,677 to 147,837 base pairs, revealed the typical quadripartite structure of flowering plant plastomes, with a large single-copy region, a small single-copy region and a pair of inverted repeats. Phylogenetic trees built from these plastome sequences placed both L. betpakdalensis and L. lasiocarpa within the same major clade, Clade Fa, together with close relatives L. macrantha, L. semiglabra, L. pavlovii and L. effusa. But here the plastome data hit a wall: most internal nodes within Clade Fa received bootstrap support below 50 percent, leaving the relationship between the two focal species unresolved. The researchers attribute this poor resolution to the recent diversification of Lappula — most extant species diverged within roughly the last 3 million years — combined with the relatively slow evolutionary rate of plastomes, which simply have not accumulated enough informative variation to separate recently split species.

The nuclear data told a sharper story. Using the Angiosperms353 probe set — a standardized panel of hundreds of low-copy nuclear genes designed for flowering plant phylogenomics — the team recovered 234 to 347 loci per sample, averaging 324 loci. After quality filtering, 274 high-quality loci were retained. Both concatenation-based and coalescent-based analyses, the latter using ASTRAL to infer a species tree from individual gene trees, produced largely congruent topologies with substantially higher support than the plastome trees. Within Clade VIIb of the nuclear phylogeny, the four samples of L. betpakdalensis formed a strongly supported monophyletic lineage — local posterior probability of 1.00 and bootstrap support of 100 percent — sister to a clade comprising L. lasiocarpa, L. macrantha, L. semiglabra and L. pavlovii. In other words, the nuclear genome sees two independent evolutionary lineages, not one variable species.

The phylogenomic results also settle a long-standing systematic question: where does L. betpakdalensis belong within Lappula? The answer is section Macranthae, as circumscribed by Ovczinnikova in 2005. The species’ nutlets, measuring 2.2 to 2.7 millimeters long with a thickness-to-length ratio of roughly 0.15, are dorsiventrally compressed — a distinctive nutlet type shared across the section. Popov had originally placed the taxon in Lepechiniella sect. Lophoptereae, characterized by coarsely dentate-winged nutlets, and Ovczinnikova later proposed sect. Kazachstanicae. But the phylogenomic evidence does not support Lepechiniella as a distinct genus at all; instead, it confirms Nabiev’s broader circumscription of Lappula, consistent with multiple recent molecular studies.

Comparisons with L. omphaloides, the species under which the taxon was first described as a variety, further reinforce the distinction. Beyond the spreading versus appressed-sericeous indumentum, the shorter pedicels, the smaller corolla and the narrower nutlet wings, the relative length of the stigma versus the nutlets — a rather conservative character within Lappula — separates the two: the stigma of L. betpakdalensis does not overtop the fruit, while that of L. omphaloides exceeds it. Individuals of L. omphaloides also bear an appressed-sericeous-gray hair type, uncommon in the genus and recorded elsewhere only in L. sericata and L. transalaica. Because material of L. omphaloides was not sampled, its precise phylogenetic position remains open for future work.

Taken together, the chloroplast and nuclear phylogenomics and the detailed morphological comparisons converge on a single conclusion: L. betpakdalensis and L. lasiocarpa are distinct species, and the 1995 synonymization is rejected. The reinstatement is more than a bookkeeping exercise. It restores a lineage with a real evolutionary history, a real geographic range — currently known only from the Lake Balkhash region of Kazakhstan — and a real set of diagnostic characters that field botanists can use. It also highlights a broader lesson for biodiversity science: as phylogenomic tools become cheaper and more powerful, species once lumped together on the basis of superficial similarity are being pulled apart, and the true diversity of Central Asia’s flora is only beginning to come into focus.

Subject of Research: Botany

Article Title: Reinstatement of Lappula betpakdalensis ( Boraginaceae ) based on phylogenomic and morphological evidence

Article References: Liu, D.-H. (2026). Reinstatement of Lappula betpakdalensis (Boraginaceae) based on phylogenomic and morphological evidence. PhytoKeys, 279, 35-49. https://doi.org/10.3897/phytokeys.279.207568

Image Credits: AI Generated

DOI: 10.3897/phytokeys.279.207568

Keywords: Lappula, Boraginaceae, phylogenomics, taxonomy, species delimitation, Central Asia, Kazakhstan, chloroplast genome, Angiosperms353, nutlet morphology, Reinstatement, betpakdalensis