Microhabitat on rocky ground: exposed sand and gravel, lichen cover on rocks and a sparse cover of low-growing plants.

In late November, we travelled to Nuwerust Farm in the eastern Cederberg, Western Cape, for field observations. Our major goal: finding Pelargonium centrifugum in the wild. This meant catching the flowering period (October-December) and photographing plants in situ, in the context of their habitat. For a recently described species with an extremely limited geographical range, this is no simple task.

All known records of P. centrifugum are from the Breekkrans river valley on the eastern side of the Cederberg mountain range. The protologue states that the species is known from only two neighbouring sites, and that populations there are small. Notably, flowering is illustrated with photographs of plants in cultivation in the Northern Hemisphere, whereas under natural conditions only a photograph of a plant with leaves is provided.

The forecast was clear: what lay ahead would not be a straight-forward search.

Eastern Cederberg, Western Cape: sandstone outcrops and rocky slopes in the Breekkrans river valley.

Hidden in plain sight

P. centrifugum is a deciduous geophyte belonging to the Hoarea section. For most of the year the plant remains hidden in the soil as an underground tuber and is almost impossible to notice; its above-ground organs appear seasonally during the active period and only for a short time.

The tuber is not merely a storage organ but a way to survive the season when above-ground growth is nearly impossible. It remains underground when conditions are hot and dry and the thin soil layer loses moisture rapidly.

Typical habitat of P. centrifugum: a rocky slope with sparse low shrub cover and large sandstone blocks.

This life cycle makes geophytes both drought-tolerant and inconvenient for the observer to locate – outside the active phase the plant cannot be detected, as it is completely concealed below ground. As a result, even with the coordinates in hand, one can walk the slope several times and still miss the plant until the flowering stalks emerge above the surface.

P. centrifugum is hysteranthous*, like most species in sect. Hoarea. During anthesis, the plant may appear almost leafless with only a few slender flowering stalks emerging directly from the ground. These stalks are easy to overlook against a background of dry herbaceous remains and last year’s stems. On a stony slope they can readily be taken for dry remnants of other vegetation – and one simply walks past.

Hysteranthous is a botanical term describing a plant in which flowers appear before the leaves. The flowering phase occurs first, while the plant has few or no leaves, and the leaves develop later in the season. This trait is often associated with plants adapted to dry or strongly seasonal climates, where flowering early may provide reproductive advantages (e.g., before shading, before drought, or after disturbance).

Learning by looking

Because the species is known from only a small number of records and localities, each independent field observation refines our understanding of its distribution and phenology.

Identification relies on floral characters, especially the unusual orientation of the five fertile stamens. The anthers face outward and dehisce extrorsely – from the centre of the flower towards the periphery – an orientation reflected in the name centrifugum. These characters are best assessed on living material, so reaching the site during the flowering season is essential.

The eastern Cederberg is a landscape where conditions change over very short distances: adjacent slopes may differ sharply in the type and structure of the substrate, in aspect, and in how long moisture is retained after rainfall. For plants, this translates into a set of specific microhabitats and environmental conditions: a rocky base, a thin soil layer and local moisture retention.

Rocky patches with boulders and soil pockets between blocks – potential sites to search for the geophyte.

The hunt begins

For a narrowly local endemic, precise localisation is crucial, and the search begins only once you are at the right point. Once the tar road ended, a long gravel road led us deeper into the rocky slopes. The field markers were clear: south-, south-west- or south-east-facing slopes and sandstone boulders.

Our search focused on microtopography between the rocks. We examined the bases of boulders and sandstone blocks, looking for cracks, fissures and small pockets where fine soil accumulates and moisture persists longer after rain – places where slender flowering stalks are found in the flowering season.

The search took longer than we had planned. But this is to be expected for a species occurring in small, localised groups: similar-looking rock exposures are easy to mistake for suitable habitat.

We systematically examined slopes with the appropriate aspect and checked microsites where shelter and some soil are retained. The structure of the vegetation cover was also a useful clue: sparse cover, a predominance of exposed rock and a local concentration of plants in fissures and at block bases.

Discovery and diagnostics

In the end, we found a single plant in a narrow fissure between sandstone blocks. It was small, only 6-7 cm tall. The tuber was partly covered by remnants of sandy soil; several slender peduncles were visible, and nearby there were dried remains of older inflorescences. Leaves, however, were absent, including even their dry remnants.

The flower was consistent with the description in its most conspicuous characters: five pale cream petals and red, feather-like markings on the two upper petals merging into a single blotch.

By the time the flower was photographed, the stamens had already fallen. This was unfortunate, considering the key diagnostic character of P. centrifugum – the unusual androecium structure and the orientation of the anthers. At this late stage of flowering the decisive character was therefore unavailable, and the identification had to rely on the set of features that remained visible and on the habitat context.

Flower of P. centrifugum (close-up): cream petals with red markings on the two upper petals; stamens had already fallen by the time of photography.

We documented the plant’s position, the absence of leaves, the petal colour and nectar guides, traces of previous inflorescences and the microsite itself – a small soil pocket between boulders, where such geophytes can persist year after year.

Photographing the plant between rocks turned out to be an additional technical task in its own right. Shooting was affected by the geometry of the boulders (dictating the viewpoint), the contrast between shade and direct light and movement of the slender peduncle due to gusts of wind. In the end, the photo series is clearly field documentation, but for rare taxa this kind of record is often more valuable than a carefully composed image: it shows the plant in its natural setting and documents key details of the microhabitat.

Sensitivity and survival

The habitats of P. centrifugum fall within the Succulent Karoo biome and are described as low shrubland communities where non-succulent shrubs (e.g., Berkheya, Felicia, Pteronia) occur together with succulent plants (Crassula, Euphorbia, Ruschia, Tylecodon).

According to published data, the corresponding vegetation types – Agter-Sederberg Shrubland and Swartruggens Quartzite Karoo – are not classified as threatened, and one of the localities is covered by a conservation agreement with the landowner.

Succulent plants rooted in crevices at the base of sandstone blocks – typical sites where fine soil and moisture are retained.

Nevertheless, for a species with an extremely narrow geographical range and small populations, vulnerability is determined less by the overall status of the biome than by the scale of its distribution. A few unfavourable seasons or local disturbance of microhabitats may be enough to impair the species’ survival prospects.

The practical conclusion for field work here is straightforward: minimise disturbance and document the plants as carefully as possible, increasing the value of observations without any impact on the site.

On the way back, we went over our search strategy: which landmarks we should have noted in advance, how to allocate time and which areas we had missed. For micro-local endemics, planning is useful, but it does not guarantee success – the decisive factor is the match between the site and the phenological stage.

That day we managed to find P. centrifugum in flower and photograph the plant in its natural habitat. The individual was small and almost inconspicuous on the slope, but the find allowed us to document both flowering and the species’ associated microsite.

Even with the limitation of an absent key diagnostic feature, the observation remains valuable for understanding phenological dynamics and the expression of morphological characters in representatives of section Hoarea.

About the author

Elena Ioganson is a popular science writer and photographer based in Bulgaria. Together with her husband, Boris Ioganson, she has made more than a dozen trips to South Africa, documenting Pelargonium and Monsonia species in their natural habitats. Email: eioganson@gmail.com.

Further reading

Strlič, M., van der Walt, J.A. & Manning, J.C. (2025). Two new species of Pelargonium sect. Hoarea (Geraniaceae) from the Core Cape Floristic Region, South Africa: P. saronense and P. centrifugum. South African Journal of Botany 186: 323–328.