Log in

Editor's Picks

Carlos collecting Quercus ×alentejana (Q. faginea × Q. pyrenaica) in northeastern Portugal for his PhD thesis © Carlos Vila-Viçosa
An interview with Portuguese oak conservationist Dr. Carlos...
Amy Byrne | Apr 19, 2024
Roderick Cameron | Apr 13, 2024
Pages from Gert's book
It was a great pleasure for me to be able to write about my...
Gert Fortgens | Feb 15, 2024

Plant Focus

Quercus crassipes acorns with inrolled cupule margin
One of the more well-known Mexican oaks in cultivation.

The Royal Oaks of Penrice Castle, Wales

Compared to many other countries the British Isles have a paltry oak flora with just two species, so they really shouldn’t cause any confusion; but they have been doing so for years, if not centuries.

James Brown, writing in 1851, put his finger on it, thus: “There are, indeed, many who would deny that there are two species of the oak found in our forests; for, say they, ‘the distinguishing characters of each do not always hold good.’ This I readily admit; for I am daily in the habit of seeing oak trees, and of examining them, and I daily see that, so far as the distinguishing characters that have been pointed out go, they are very often blended together in one tree. But this is easily accounted for by the two species having been long growing in the country together, from which have arisen many trees, which are no doubt what may properly be termed hybrids between the two.” Well, possibly.

A royal oak at Penrice Castle
One of the Royal Oaks at Penrice Castle, Wales, planted by King George V on 18 July 1920 © Sairus Patel

Our two natives are the pedunculate oak, Quercus robur, and the sessile or durmast oak, Q. petraea. A hybrid between them, occurring on the European Continent, had been named by Bechstein as Q. ×rosacea as early as 1810. The suggestion that the identification of the two species is often made difficult by hybridization has been reiterated regularly down the years, and Clive Stace’s authoritative New Flora of the British Isles (Fourth Edition, 2019) states that Q. ×rosacea “occurs throughout BI (British Isles) in areas where 1 or both parents occur, occasionally being commoner than either. It combines in various ways the leaf and fruit characters of the 2 spp., and is fertile.”

It has occasionally been argued that one of the parent species is not truly native. In his A History of the Vegetable Kingdom (1866) William Rhind declared that: “The species [Q. petraea] is supposed to have been introduced, some two to three ages ago, from the continent, where the oaks are chiefly of this latter species, especially in the German forests, the timber of which is known to be very worthless. But what is of more importance to us is that, de facto, the impostor abounds and is propagated in the New Forest and in other parts of Hampshire, in Norfolk and in the northern counties, and about London.” Sorry, Germany.1

Mark Anderson, Professor of Forestry at Edinburgh University, identified the alternative impostor (1950): “There is such a mass of evidence pointing to this typically Continental species [Q. robur] not being native to these islands, certainly not of Ireland and Scotland, that I do not hesitate to classify it as introduced.”

Participants at the Oak Open Day at Penrice discuss one of the Royal Oaks at Penrice
Participants at the International Oak Society Oak Open Day event at Penrice Castle discuss the oak planted by King George V in 1920 © Sairus Patel

Professor Anderson was one of the minority of workers in this field who did not accept that hybrids were a common occurrence. His view was supported by Eustace Jones of the Oxford Forestry School. He was the author (1959) of the monograph of the genus Quercus for the Biological Flora of the British Isles, and was an authority to be reckoned with. He noted that attempts by Dengler (1941) to produce hybrids by cross-pollination had only been successful in about two percent of cases, rising to about 15 percent on one occasion. He concluded that most descriptions of hybrids resulted from a misunderstanding of the natural characters of each species, which were far more variable than had hitherto been assumed. There is an irony to this, because his highly accurate descriptions of the leaf and fruit morphology of the two species would later form the basis of work by several investigators (Carlisle and Brown 1965, Cousens 1962, 1963, and 1965, Wigston 1974 and 1975, and Potter 1990) that would demonstrate that in this opinion he was mistaken: it’s now generally accepted that both species are native, that they can hybridize and that some of the variation exhibited by them is the result of introgressive hybridization.

A royal oak showing a long peduncle, a Q. robur characteristic
Detail of a Royal Oak at Penrice, referred to in the analysis below as Quercus ×rosacea aff. robur, with short petioles and long acorn stalks characteristic of Q. robur, but shallow, regular lobing typical of Q. petraea © Sairus Patel

It’s known that the two parent species produce hybrid F1 offspring very rarely. Those F1s may or may not be sterile with respect to each other, but they are fertile with respect to either of the parent species, and in any case the occasional F1 will be greatly outnumbered by—and thus be far more likely to pollinate or be pollinated by—pure sessile or pedunculate oak. Hence the F2 generation will be back-crosses with the parents, and we can expect the leaf morphology to show roughly 25% of the characteristics of one species, and 75% of the other. With further back-crossing we will see a phenotypic “drift” towards the character of one of the parents, the direction of that drift being determined by natural selection in the wild, or very possibly by human selection in parklands and gardens.

And so to the parkland of Penrice Castle on the Gower Peninsula in southwest Wales. At the kind invitation of Thomas Methuen-Campbell, members of the IOS gathered in early July of this year to inspect the impressive collection of oaks. The land has been in the care of the same family for 29 generations since the twelfth century (more information is available here), and royalty have been frequent visitors. In the first three decades of the last century it became customary for visiting members of the Royal Family to plant an oak in the parkland just to the west of the house. One of these is a Turkey oak, Q. cerris L., and the other eight are natives: but of which species? Conversation soon turned to questions of ancestry and breeding.

Miss Emily Talbot planting royal oak at Penrice
Miss Emily Talbot, at the time owner of Penrice Castle, at the planting of one of two oaks by Edward VII, 1904; image courtesy Thomas Methuen-Campbell

Some years ago I had developed a means of identifying the two species and their range of hybrid forms based on leaf morphology (Potter 1994). A description can be viewed here. Thomas and I determined to use this method to investigate those questions of ancestry and breeding. We collected at least 15 leaves from the lower crown of each of the trees, and I assessed them for the length of the petiole as a proportion of the length of the lamina (P); the average number of leaf lobe pairs (L); and the depth and regularity of lobing, the presence or absence of auricles at the leaf base, and the presence or absence of abaxial leaf hairs (combined to give a total leaf character score, C). This results in a three-dimensional cloud of data points that is most easily viewed as two elevations: L against C and P against C, Figure 1.

Figure 1
Figure 1

From this we can see that tree number 6 (planted by the late Queen Mother, then Duchess of York, in 1926) conforms to the archetype of Q. petraea as described by Eustace Jones, op. cit. Tree number 8, planted in 1938, conforms to Q. robur. The remaining six trees are hybrids. It might be thought at first glance that tree number 7 (planted in 1936) is also Q. robur, but reference to the P/C elevation on the right shows that the petioles are slightly too long to qualify.

The rules of taxonomy require that these six trees should be referred to Q. ×rosacea. I would respectfully suggest that a single name is insufficient to reflect the full range of variation and so for practical purposes I would refer to tree number 4 as Q. ×rosacea aff. petraea, and to numbers 1, 2, 3, 5, and 7 as Q. ×rosacea aff. robur.

The Royal Oaks at Penrice Castle

  • 1904: two oaks planted by Edward VII and Emily Talbot (numbers 1 and 2 in Figure 1)
  • 1920: oak planted by George V (number 3)
  • 1923: oak planted by the Prince of Wales, later Edward VIII (number 4)
  • 1926: two oaks planted by the Duke and Duchess of York, later George VI and the Queen Mother (numbers 5 and 6)
  • 1936: no further details available (number 7)
  • 1938: oak planted by Princess Mary (number 8)


Miss Emily Talbot participates in the planting of royal oak 1904
Miss Emily Talbot helps plant an oak (number 1 or 2, Quercus ×rosacea aff. robur), 1904; image courtesy Thomas Methuen-Campbell

Both of the native species, and their hybrid, have been recorded from the Gower Peninsula (National Biodiversity Network Atlas Wales, available to view here). Much of the peninsula is underlain by Carboniferous limestone, but there is a band of Namurian sandstone running down the valley on which Penrice Castle stands. This would have given rise to freely draining acidic loamy soils, on which we might expect Q. petraea to be the dominant oak. As Q. robur is also present in the area we might expect some introgression to have occurred and if unfettered natural selection had been operating we would expect a drift toward the Q. petraea parent. But the Royal Oaks show a drift in the opposite direction, toward a Q. robur parent. What’s been going on?

There are no records of the origins of these oaks: we don’t know whether they were raised from acorns collected on the estate or whether they were bought in as plants. It’s unlikely that young plants would have been marketed as Q. ×rosacea, so that points somewhat in the direction of local acorns rather than a commercial nursery, but even if they were bought in, they most likely would have been of a local provenance. I believe that Eustace Jones provides an explanation for the pattern that we see (op. cit., p. 175): “It is very difficult to disentangle the part played by planting from that played by soil and climate in determining the distribution of the two species. Q. petraea has often been replaced by Q. robur not only because of a belief that the timber of the latter was superior . . . but because its more frequent fruiting, its larger and more easily stored acorns, and its more vigorous seedling naturally lead, in the absence of any conscious preference for Q. petraea, to the predominance of Q. robur in nurseries.”

So it seems that introgressive hybridization has been at work on the oaks in and possibly around the grounds of Penrice Castle, but that natural selection which would have led to a phenotypic drift in the direction of Q. petraea has been overridden by human selection favoring a Q. robur phenotype. This presents us with an intriguing opportunity. It’s possible that future generations will want to preserve the legacy of the Royal Oaks by collecting acorns from them and planting their seedlings. If records of the parentage of those seedlings were maintained, and similar assessments of their leaf morphology undertaken and recorded, the further direction of phenotypic drift could be determined. There’s a good chance that our descendants could witness microevolution in action.

Edward VII looks on
Edward VII looks on as an oak is prepared for planting, 1904; image courtesy Thomas Methuen-Campbell


My grateful thanks to Thomas Methuen-Campbell, the current custodian of Penrice Castle and its Royal Oaks, for hosting the original IOS visit, and for inviting me to carry out the analysis which resulted in this article.

Postcard of Royal Oak
Postcard showing one of the oaks planted by Edward VII; source eBid

Works cited

Anderson, M.L. 1950. The Selection of Tree Species. Edinburgh: Oliver and Boyd.

Brown, J. 1851. The Forester: a Practical Treatise on the Planting, Rearing and General Management of Forest Trees. Edinburgh: W. Blackwood and Sons.

Carlisle, A., and Brown, A.H.F. 1965. The assessment of the taxonomic status of mixed oak (Quercus spp.) populations. Watsonia 6 (2): 120–127.

Cousens, J.E. 1962.Notes on the status of Sessile and Pedunculate oaks in Scotland and their identification. Scottish Forestry 16: 170–179.

Cousens, J.E. 1963. Variation of some diagnostic characters of Sessile and Pedunculate oaks and their hybrids in Scotland. Watsonia 5: 273–286.

Cousens, J.E. The status of pedunculate and sessile oaks in Britain. Watsonia 6: 161–176.

Dengler, A. 1941. Bericht über Kreuzungsversuche zwischen Trauben- und Stieleiche und zwischen europäischer und japanischer Lärche. Mitteilungen der Hermann-Göring-Akademie der Deutschen Forstwissenschaft 1: 87–109.

Jones, E.W. 1959. Biological Flora of the British Isles: Quercus L. Journal of Ecology 47: 169–222.

National Biodiversity Network Atlas Wales. 2023. https://wales.nbnatlas.org/ Accessed 26 July 2023.

Potter, S.M. 1990. A method for the identification of the native British oaks and their intermediate forms, and its application to field studies. Unpubl.

Potter, S.M. 1994. Practical methods for distinguishing the two native British oaks and their intermediates. Quarterly Journal of Forestry 88(1): 27–34.

Rhind, W. 1866. A History of the Vegetable Kingdom. London: Blackie and Son.

Stace, C. 2019. New Flora of the British Isles, Fourth Edition. C&M Floristics.

Wigston, D.L. 1974. Cytology and genetics of oaks. In Morris, M.G. and Perring, F.H. (edit.) The British Oak: its History and Natural History: 27–50. London: the Botanical Society of the British Isles.

Wigston, D.L. 1975. The distribution of Quercus robur L., Q. petraea (Matt.) Liebl. and their hybrids in south-western England. 1. The assessment of the taxonomic status of populations from leaf characters. Watsonia 10: 345–369.

1 I have heard it claimed that the curved timbers that would form the futtocks ("foot-oaks") and ribs that determined the shape of the hull of a British naval man o’ war were usually cut in a single piece from hedgerow trees, which were likely to be Q. robur, whereas Continental ship builders used forest-grown Q. petraea which, being straighter, had to be jointed. It was claimed that this made British ships stronger and better able to withstand a broadside than French or German vessels. It may be that this gave rise to the erroneous belief that the timber of the former species was superior to the latter. It is an early example of British jingoism being applied to silviculture.