Aston Rowant Roses

Geoff Jones examining Rosa rubiginosa on Beacon Hill against the backdrop of the Oxfordshire plain

Late summer and early autumn is the best time to identify wild roses if one is game for the challenge. While it is a challenge, there is so much we don't know about the status and distribution of Rosa taxa in the county that their study really pays off I have found (though I am by no means an expert). I believe Geoff Jones, new to roses and a volunteer recorder at Aston Rowant National Nature Reserve, whom I met there on Saturday to spend a few hours rhodologising, has come to be of the same opinion. I will always be delighted to hear from botanists such as Geoff who are intrigued by these beautiful plants and would like to record them thoroughly. I hope the following account of our short visit will give you an idea of what could be out there to be found in your patch.

Aston Rowant NNR is of course well-known for its chalk grassland and woodlands, but it is also a very scrubby place and this is where to look for roses. Some of the scrub is very old and these are the best places to hunt for unusual roses. We started on Beacon Hill where Geoff had already puzzled over a few potentially interesting roses. In scrubby bits of downland we found several grotty bushes of two sweet-briars, Rosa micrantha (small-flowered) and R. rubiginosa (sweet-briar or Eglantine). The UK has three native species of sweet-briar — the third, R. agrestis (small-leaved), was only rediscovered in the county last year at nearby Pyrton Hill.

Sweet-briars readily colonise under-grazed grassland, R. rubiginosa chalk grassland especially. They can be told from other roses by the dense covering of stalkless sticky glands on both surfaces of the leaflets that give off an apple-like aroma when rubbed. Bright shiny red when fresh, as the glands dry up (as in hot weather or in the herbarium) they form an obvious white, speckley crust on the leaves. In these two species, the pedicels and hypanthium (and hips) are also covered in glands, borne on eye-catching bristle-like stalks. R. rubiginosa is the smelliest of the three, detectable by nose from a few feet away on a hot day, and it has the fiercest armature, with densely prickly stems with both stout hooked prickles and fine needle-like prickles (acicles). The leaves of both R. rubiginosa and R. micrantha are round with neat multi-serrated margins, but R. rubiginosa has larger flowers that are a deep pink colour and the sepals are held erect from the hips, pointing straight out the top.

Rosa rubiginosa showing glands over the leaves and hips and the erect sepals. The glands are visible as shiny white spots (click to enlarge).

An observation from my examination of roses elsewhere that we were able to confirm at Beacon Hill was the frequent occurrence of the hybrid between R. canina and R. caesia subsp. vosagiaca (glaucous northern dog-rose). R. caesia does not occur in our area but this hybrid, which goes under the name R. x dumalis, is widespread across southern England and in several counties is one of the commonest roses. This certainly seems to be my experience in Oxfordshire, where it seems to behave very much like a species and is locally common. R. caesia subsp. vosagiaca x canina is quite a straightforward rose to identify on account of its considerable vigour, often to be seen with stout wine-red young shoots waving out the tops of uncut hedgerows and bursting with a profusion of clusters of big hips in the autumn.

Rosa x dumalis (click to enlarge): dull greyish leaflets that remain folded along their lengths as they age; purple-tinted young stems becoming deeply-stained, thick and stoutly-armed as they age; hips usually large, elongated and borne in tight clusters, the bracts very leafy and broad; the styles emerge from the orifice in a low dome, wider than that in R. canina; the prickles are more variable but are often squat, stout and well-hooked.

Geoff also took me to look at the roses in the scrub along an old holloway on the other side of the M40 from Beacon Hill where he'd found some downy roses. On our way down we spotted a rose with very conical tops to the hips. Occasionally one finds dog-roses with slightly convex or mamillate tops to the hips (the disc), but truly conical discs are a feature only of R. stylosa (short-styled field-rose) and hybrids therewith. R. stylosa is rare in Oxfordshire, with only two post-2000 sites, but our bush was not the real deal but its hybrid with R. canina, known as R. x andegavensis. I hadn't come across any roses like this outside my patch where it is occasional around Otmoor. It is quite a common taxon in other vice counties where R. stylosa is also rare. The range of variation within R. x andegavensis is great, and plants can look only subtly different from R. stylosa, but the bush at Aston Rowant was closer to R. canina with only the disc and styles showing any strong evidence of R. stylosa.

Cross-section of hip of Rosa x andegavensis with strikingly conical disc and styles emerging in a short narrow column, features of R. stylosa.

The downy roses Geoff had found were indeed interesting. Like sweet-briars downy roses are also very glandular on the leaves and pedicels but they have a thicker indumentum of longer hairs on the leaves so that they appear soft and dull. The leaves are oval rather than round, and the glands are smaller and pale, smelling only weakly resinous, and the prickles are distinctively straight and patent. There seemed to be two taxa along the lane: one with long pedicels (over 2cm), round hips and patent sepals (R. tomentosa (harsh downy rose) characters); the other with bigger hips and erect to sub-erect sepals, and some smaller slenderer prickles among longer arcuate ones (R. sherardii (Sherard's downy rose) characters). As both lacked glands among the hairs on the leaves these could not be pure downy roses but hybrids with R. canina, which being a rather featureless rose has the effect of diluting characters of other species, making our plants (in my view) R. x scabriuscula and R. x rothschildii, respectively. R. sherardii is another northern rose that is very rare in the south, but Druce and other older botanists knew it from quite a number of locations in Oxfordshire, including nearby Crowell. Perhaps it was was once a little more common but has been lost to scrub/hedgerow clearance and hybridisation. In my experience R. x scabriuscula is also commoner than its parent R. tomentosa.

The above wasn't all we found! There were also plain old dog-roses you'll be glad to hear, but even to identify these one needs to eliminate hybrids and understand the variation within R. canina — we found the commonest two variants, the hairy and intermediate forms. There were also several bushes showing features of R. micrantha and R. obtusifolia (round-leaved dog-rose), a dog-rose that is not uncommon around Otmoor and whose hybrid with R. canina (R. x dumetorum) I am finding to be one of our commonest roses. I have found possible R. micrantha x obtusifolia at one other site but it is otherwise not known from the county and is rare nationally, and so needs confirming by the BSBI roses referee. In all then, we found eight, perhaps nine taxa, good going for threes hour's work:

The long glandular-setose pedicels, globose hips and patent sepals of Rosa tomentosa but on a plant without glands on the hairy leaves, indicating that it is a hybrid with R. canina

• Rosa arvensis (field rose)
• Rosa canina group 'Transitoriae'
• Rosa canina group 'Pubescentes'
• Rosa caesia subsp. vosagiaca x canina (=R. x dumalis)
• Rosa canina (female) x stylosa (male) (=R. x andaegavensis)
• Rosa micrantha
• ? Rosa micrantha (female) x obtusifolia (male)
• Rosa rubiginosa
• Rosa tomentosa (female) x canina (male) (=R. x scabriuscula)
• Rosa sherardii (female) x canina (male) (=R. x rothschildii)

Many thanks to Geoff for his company and for showing me what he has been working on. I don't doubt he'll continue to find interesting roses at Aston Rowant.