Hydrangea mathewsii Briq. — Rusty sprawling hydrangea

Introduction

Rusty sprawling hydrangea is a functionally dioecious Peruvian root-climbing liana distinguished by an unusually textured combination of foliage and indumentum. Its inflorescences are generally branched or thyrse-like, with white reproductive flowers and no enlarged marginal flowers. The lower leaf surface is conspicuously rugose, with dense stellate hairs concentrated along the veins, while the branchlets and inflorescence branches carry reddish stellate hairs. These characters give the plant a distinctly rougher and rustier appearance than the closely entangled H. jelskii (McClintock 1957; Sutton 2025).

The species was described by John Briquet in 1919 from Peruvian material collected by Andrew Mathews. Surviving type material includes Mathews 3051 specimens at Kew and Edinburgh. Unlike many Cornidia names that accumulated large synonym complexes during the twentieth century, no formal botanical synonym was retained for H. mathewsii in the modern audit; the problem here is not nomenclatural proliferation but determining where the biological limits of the species actually lie (World Flora Online Consortium 2026).

Modern field work confirms that the taxon is not merely an old herbarium concept. Material identified as H. mathewsii was recollected near Leymebamba in Chachapoyas, northern Peru, during the extensive Cornidia field program that underpinned the molecular work. That geographic overlap with intensively sampled H. jelskii populations is particularly important because the two taxa subsequently proved closely interwoven genetically (Granados Mendoza 2013).

Granados Mendoza et al. (2021) recovered H. mathewsii inside Clade F, the same Andean assemblage containing H. jelskii, H. tarapotensis, and two unnamed morphotypes. No exclusive H. mathewsii lineage emerged. Individual H. mathewsii accessions were repeatedly paired or grouped with H. jelskii, and another was associated with Hydrangea sp. 3. In other words, the two named species intermingle at the level of the sampled nuclear loci even though their characteristic foliage remains distinguishable.

That combination of evidence is more informative than either morphology or the gene tree alone. The rugose, densely vein-hairy lower leaf surface and reddish indumentum provide a repeatable organismal signal for H. mathewsii, while the nuclear results show that the lineage has not achieved—or has not retained—complete genealogical separation from neighboring forms. Recent divergence, ancestral polymorphism, introgression, or hidden subdivision within one or both currently named species could all generate such a pattern.

The unidentified sp. 3 and sp. 4 entities make premature lumping especially undesirable. If H. jelskii and H. mathewsii were simply collapsed because they intermingle in two genes, the resulting broader species would still contain morphologically recognizable unnamed lineages and would tell readers less rather than more about Andean Cornidia. A type-anchored genomic study should instead ask whether the current names correspond to genomic clusters, ecological populations, or parts of a reticulate species network.

This makes H. mathewsii one of the better examples in Tree TSAR of a species retained not because its boundaries are considered finished, but because its present morphology is sufficiently coherent that removing the name would outrun the evidence. The species should therefore be presented as accepted with an explicit species-complex flag, not as either unquestionably isolated or merely synonymous with H. jelskii.

Horticultural and Agricultural Uses

No reliable ornamental or agricultural use of H. mathewsii is documented. Its rust-colored stellate indumentum and strongly textured lower leaf surface could make authenticated material visually distinctive, but horticultural promise should not be substituted for actual cultivation records.

Botanical cultivation would be especially valuable for comparative purposes. Growing H. mathewsii beside geographically matched H. jelskii would permit direct assessment of juvenile and adult foliage, pubescence development, whole-plant architecture, flowering phenology, and sexual dimorphism—the exact classes of characters that are difficult to compare from disconnected herbarium collections.

Collections should be provenance-rich. A single clone labelled H. mathewsii could preserve neither the geographic variation within the species nor the information needed to decide whether the present Peruvian concept hides additional lineages.

Conservation Concerns

No formal IUCN Red List assessment was located for H. mathewsii. POWO currently recognizes it as a Peruvian endemic of the wet tropical biome, which makes its geographic scope appreciably narrower than the nominal range of H. jelskii (Royal Botanic Gardens, Kew 2026).

That apparent endemism deserves attention, but an independent threat category would be premature without better knowledge of population number and distribution. Conversely, the absence of an assessment should not be read as evidence of security. Cornidia species are disproportionately associated with humid primary and montane forest, and several members have become much rarer once old broad species concepts were dismantled (Sutton 2025).

The unresolved boundary with H. jelskii also creates a practical conservation problem. If H. mathewsii contains multiple Peruvian lineages, conserving only one named accession would underrepresent the species. Genomic work should therefore be designed simultaneously as a species-delimitation and conservation-provenance study.

Additional Information

Plants of the World Online — Hydrangea mathewsii (opens in a new tab): recognizes the species as endemic to Peru.

World Flora Online — Hydrangea mathewsii (opens in a new tab): provides nomenclatural information and links to surviving Mathews type specimens.

Trees and Shrubs Online — Section Cornidia (opens in a new tab): gives the most useful concise modern diagnosis of its rugose, densely hairy foliage.

iNaturalist observations — Hydrangea mathewsii (opens in a new tab): may help locate photographic records, although identification against H. jelskii should be treated cautiously.

References and Further Reading

Briquet JI (1919) Decades plantarum novarum vel minus cognitarum: Decades 17–25. Annuaire du Conservatoire et du jardin botaniques de Genève 20: 342–427. https://www.biodiversitylibrary.org/item/112421 (opens in a new tab)

Granados Mendoza C (2013) Tree top climbing: Towards an evolutionary tree of Hydrangea section Cornidia. PhD thesis, Faculty of Sciences, Ghent University, Ghent, Belgium. https://biblio.ugent.be/publication/4211473 (opens in a new tab)

Granados Mendoza C, Martínez Salas EM, Goetghebeur P, Wanke S, Samain M-S (2021) Molecular phylogeny, character evolution, and biogeography of Hydrangea section Cornidia, Hydrangeaceae. Frontiers in Plant Science 12: 661522. https://doi.org/10.3389/fpls.2021.661522 (opens in a new tab)

McClintock E (1957) A monograph of the genus Hydrangea. Proceedings of the California Academy of Sciences, Fourth Series 29(5): 147–256. https://www.biodiversitylibrary.org/part/51707 (opens in a new tab) https://doi.org/10.5281/zenodo.15905786 (opens in a new tab)

Sutton J (2025) Hydrangea species of Section Cornidia. Trees and Shrubs Online. International Dendrology Society. https://www.treesandshrubsonline.org/articles/hydrangea/hydrangea-species-of-section-cornidia/ (opens in a new tab)