Rhododendron ponticum: a problem plant, ill-served by misinformation and ignorance.
Rhododendron ponticum, a mauve-flowered European native rhododendron, was introduced to the British Isles in the 18th century and planted widely as a colourful evergreen. It is now an out of control weed in western parts of the UK and is on the verge of being proscribed in both National and EC-wide Legislation so that it can no longer be sold or planted. I support this ban being applied in the UK as this plant is clearly out of control in some areas and in addition, R. ponticum is a vector for the spread of Phytophthora ramorum into forestry plantations.
In 2011 Dr James Cullen wrote a paper Hanburyana (Vol 5 June 2011) claiming that wild/naturalised Rhododendron ponticum in the UK is a hybridised superweed which he describes as R. x superponticum.
I read it at the time and found it unconvincing. I was also aware that Dr Richard Milne, based in Edinburgh, had already written a paper on this same subject, based on his study of the DNA of 260 samples of naturalised R. ponticum. James Cullen should have referred to the Milne paper in detail in his own paper as it covers with rigorous science what James Cullen does with outdated visual morphological study.
My belief is that Cullen’s paper and his ‘new’ species is based on unsound data, poor observation and minimal science and we should ignore its findings. Disappointingly the article has been widely quoted in the press and on line but not hitherto critically evaluated.
R. ponticum in upper Loch Torridon, Scotland
R. ponticum occurs in its wild habitats in isolated parts of Spain and Portugal and in larger populations in the mountains of E. Turkey, Georgia and adjoining countries. These very disjunct populations are probably the result of climate change following the last Ice Age. Fossil records suggest that R. ponticum was once widespread in Europe including western Ireland (Jessen, 1948) but that extreme cold and dry lowland conditions probably resulted in its extinction over most of Europe. It might well have (re-)established as a British native had the landbridge not been submerged creating the British Isles.
R. ponticum, one of 900 species of rhododendron,is the only rhododendron species or hybrid which is a major invasive weed in the west of the UK and parts of Brittany, France. Rhododendron luteum, the scented yellow azalea species has naturalised in some areas but is not causing anything like the environmental problems that R. ponticum has. Some west coast woodland gardens in the UK such as Stonefield in Argyll have self-seeding of many species and hybrid rhododendrons but these have not escaped the bounderies of the garden and naturalised in the countryside. Ledum palustre (now Rhododendron tomentosum) may be native to the UK and can be found in bogs in central Scotland while North American Ledum groenlandicum (now Rhododendron groenlandicum) is naturalised in parts of the peak district.
R. ponticum has the ability to spread by seed or layers and can regrow from roots or small parts of stems so it hard to eradicate with physical or chemical means. In areas with high rainfall and acidic soil it spreads rapidly and can form a virtual monoculture. In recent years it has become susceptible to Phythopthora ramorum and P. kernoviae, particularly in combination with Larix kaempferi plantations and this had led to greater efforts to eradicate it.
In October 2014 I went to study one of the best known wild populations on the mountain of Foia (902m), Monchique in the Algarve in Portugal. On the north side of the mountain there are substantial populations of R. ponticum growing in open hillsides with Cistus and Rubus. The population was flowering freely, producing plenty of seed and young seedlings were observed. In every way, morphologically this population matches the R. ponticum I have observed in many sites in the UK from Cornwall to North West Scotland.
Visible characters include:
1. Dark green narrow pointed leaves,
2. leaf retention for many years,
3. often red stems,
4. narrow curved seed capsules.
R. ponticum on Foia, Monchique, S Portugal, (902 m) October 2014 showing the typical narrow deep green leaves and red stems.
In 1962 my father Peter Cox and Peter Hutchison collected R. ponticum in Turkey. (AC&H 205) These collections have proved to be quite distinct morphologically from the R. ponticum from the Iberian peninsula and thus far have not shown any propensity to seed, layer or spread from where it was planted at Glendoick over 50 years ago. It is possible that if allowed to cross with Iberian forms, evolve further and be equally invasive but again there are no examples of this at Glendoick. Dr Richard Milne and Fatima Sales observed significant variation in the Turkish R. ponticum in their field studies and they concluded that we only have a small part of the wild gene pool of R. ponticum in cultivation.
R. ponticum AC&H from the Caucasus.
Milne & Abbott (2000) examined samples of DNA of naturalised R. ponticum and found that that all British and Irish naturalised R. ponticum is derived from Iberian stocks and not from Caucasus provenance. The Iberian form of R. ponticum has been described as R. ponticum ssp. baeticum to distinguish it from the Caucasus forms. The Edinburgh revision of Genus Rhododendron part authored by James Cullen did not recognise this subspecies. As the population in the Iberian peninsula is morphologically distinct and the disjunct populations are hundreds of miles apart, it would be sensible to maintain this subspecies. This should be the name used to describe the UK naturalised R. ponticum and not R. x superponticum. The Caucasus forms would be known as R. ponticum ssp. ponticum.
R. ponticum and several north American rhododendron species including R. catawbiense and R. maximum were introduced in the 18th century to the UK. Almost immediately, nurserymen began crossing the species and creating hybrids which were planted in gardens all over the UK.
James Cullen was an ‘old school’ taxonomist who did not use DNA as a tool to his work, despite its availability. Cullen therefore relies entirely on morphological characters (i.e. characters than can be seen). He lists errant morphological characters observed in some of the sample material gathered as R. ponticum for his study from many locations in the UK and suggests putative parent species:
1. Pedicels with dendroid hairs: 4 specimens. This is a character of R. catawbiense.
2. Calyx more than 3mm: 21 specimens. This is a possible character from R. maximum.
3. Ovary rounded, hairy (generally red or brown tomentose): 5 specimens This is a character of R. catawbiense, R. maximum (and R. macrophyllum.)
4. Ovary rounded, hairy and glandular: 2 specimens. This is a character of R. maximum.
5. Leaf length/breadth ratio less than 2.5, the apex and base rather rounded: 3 specimens. A character of R. catawbiense’
Dr Richard Milne, based at University of Edinburgh sampled the DNA of R. ponticum from samples taken from many sites in the British Isles in his 2000 study. His conclusions were:
1) There are genetic markers for R. catawbiense spread across many British R. ponticum populations at low frequency, and that these are more common in colder regions.
2) The hairy ovary found in some specimens of UK R. ponticum, suggests some R. catawbiense involvement.
3) R. maximum appears to transmit occasional long calyx lobes to some populations of naturalised R. ponticum but these are not common.
3) Red corolla flecks in some specimens suggest involvement of a species outside subsection Pontica. This is probably R. arboreum, introduced in the 18th century from India and crossed many times with R. ponticum to create swarms of hybrids known as ‘Altaclarense’, ‘Cornish Red’ and ‘Sir Robert Peel’ These hybrids are common in Argyll, Ireland and Cornwall in old gardens where they form the largest rhododendron bushes known in the UK.
So let’s accept therefore that some North American rhododendron species genes are involved in small quantities in some naturalised UK R. ponticum populations. The key here is to ascertain whether the R. maximum or R. catawbiense genes would make R. ponticum more or less invasive? We might be tempted to assume ‘hybrid vigour’ but there is no evidence for this.
My father Peter Cox and I have seen little sign of these hybrid morphological characters in the many naturalised R. ponticum swarms that we have examined in Scotland, Ireland and England but if you look hard enough you can find occasional probable hybrid traits such as hairs on the ovary, suggesting some North American genes as both Cullen and Milne have stated. But from what we have observed the most aggressively invasive R. ponticum seems to exactly match the Iberian wild populations. Most of them look like 100% R. ponticum with no other species involved. The relatively small number of hybrid plants we have observed in such populations tend to stand out and show no signs of being more aggressively invasive.
Pure Iberian R. ponticum appears to be perfectly hardy in the UK climate. R. ponticum regenerates freely from damaged stems, and roots and is therefore more than capable of withstanding the effects of most cold winters in the UK, particularly in the last 40 years where winters have been milder. I observed no more than superficial foliage damage to pure R. ponticum in the 2010-11 winters which saw record low temperatures for much of the British Isles for long periods. R. catawbiense is a very hardy species so it would add extra hardiness to R. ponticum. However R. catawbiense is poorly adapted to the UK climate and does not grow easily or well. The same applies to R. maximum. Both these North American species have evolved to withstand the hot summer and very cold winters of a continental climate in E. USA and in the UK have weak root systems, chlorotic foliage and little potential as garden plants. They would not confer any genetic or evolutionary advantage to a species which already ideally suited to the climate of the British Isles.
Dr Richard Milne adds weight to this conculsion in his findings: (personal correspondence). Milne believes naturalised R. ponticum shows some evidence of introgression, through hybridisation and repeated backcrossing to R. ponticum, from R. catawbiense. This could create a situation where some British R. ponticum populations have on average something like 5% R. catawbiense in their genetic makeup. Such plants would look very like pure R. ponticum but show the occasional morphological or molecular markers of R. catawbiense. Introgression is a well known process in many wild plant populations and is common in wild populations of rhododendron species. Dr Richard Milne points out that natural selection during the introgression process would probably remove genes with a negative effect on fitness, so that after generations of seeding and maturing, eventually the only R. catawbiense genes left behind would be those with no fitness effect, or those that might confer an advantage. He suggests that such genes could lead to enhanced cold tolerance, but that we have as yet no evidence for this. It is possible that R. ponticum may evolve some hybrid vigour via introgression, but as it seems so successful an invader already, there may be no evolutionary need to do so. Cullen makes the mistake of observing populations where some hybrid characters are apparent in some populations and then assuming that all R. ponticum naturalised in the UK should therefore be regarded as hybrids. There is simply no scientific evidence for this. It is at best a hunch. For which he provides little or no evidence.
Another angle is to look at garden-raised/man-made hybrids of R. ponticum to see if they show any invasive tendencies. There are numerous man-made hybrids containing R. ponticum crossed with North American species. Most of the hybrids in purple and lavender shades are this mixture. Cultivars such as 'Fastuosum Flore Pleno', ‘Blue Peter’, ‘Old Port’, ‘Lee’s Dark Purple’ and many more. None of these shows the slightest tendency to invasiveness by continuous layering or self seeding as R. ponticum does. This is not surprising, given that North American species are not well suited to cultivation in the UK, tending to chlorosis and lack of vigour. My observation is that far from crossing with R. ponticum to make it more vigorous and invasive, the reverse is true: adding genes from these species seems to bring R. ponticum under control. In other words hybridisation into R. ponticum of any of these species makes it less invasive rather than more invasive. So the epithet R. x superponticum is therefore a misnomer.
The publication of the name R. x superponticum is not appropriate or acceptable.
What James Cullen has done is to take a naturalised population of R. ponticum with some evidence of introgression given it taxonomic status. This is not sensible or accepted taxonomic practice. In the 19th and early 20th century it was accepted practice to use the form R. x latinised name for a hybrid (usually between two defined species) or for a natural hybrid or population of natural hybrids found in the wild. R. x altaclarense for R. ponticum x arboreum for example of a latinised name of a man-made hybrid. This naming practice was ended for cultivated rhododendrons after 1950 and has not been accepted since.
When a taxon is described, it is necessary to define which characters differentiate it from its nearest relative(s). But in the case of R. x superponticum the taxon described matches R. ponticum ssp. baeticum in every consistent character; As Dr Richard Milne comments (personal correspondence) ‘there is no morphological trait that consistently distinguishes British R. ponticum from native Iberian material; instead we have occasional traits like hairy ovaries and long calyx lobes that are different, but as not all individuals have them, it cannot form the basis of a formal description.’
There is therefore simply no basis for accepting R. x superponticum as a valid taxon and I therefore propose that it be reduced in synonomy with R. ponticum ssp. baeticum or better still, just abandoned. As has been discussed, there is so far no evidence that any hybridity in R. ponticum populations make it more invasive. As Dr Richard Milne pointed out to me, even if such evidence were found, it would still not justify a new taxon name, because consistent distinguishing characters do not exist. The Cullen paper is responsible for a great deal of misinformation (you’ll find lots of articles which refer to it on the internet) and it was featured by James Wong on BBC television in The Great British Garden Revival in January 2015. I’m hopeful that this paper may go some way to redressing the balance and lets just put the ‘superponticum’ idea to bed once and for all.
Cullen, Dr James ‘Naturalised rhododendrons widespread in Great Britain and Ireland’ Hanburyana (Vol 5 June 2011)
Jessen K (1948) Rhododendron ponticum in the Irish interglacial flora . Irish Naturalists' Journal, 9, 174 175.
Milne, R. I. & Abbott, R. J. Origin and evolution of invasive naturalised material of Rhododendron ponticum L. in the British Isles May 2000: Molecular Ecology. 9, p. 541-556
The publication of the name R. x superponticum is not appropriate or acceptable.
What James Cullen has done is to take a naturalised population of R. ponticum with some evidence of introgression given it taxonomic status. This is not sensible or accepted taxonomic practice. In the 19th and early 20th century it was accepted practice to use the form R. x latinised name for a hybrid (usually between two defined species) or for a natural hybrid or population of natural hybrids found in the wild. R. x altaclarense for R. ponticum x arboreum for example of a latinised name of a man-made hybrid. This naming practice was ended for cultivated rhododendrons after 1950 and has not been accepted since.
When a taxon is described, it is necessary to define which characters differentiate it from its nearest relative(s). But in the case of R. x superponticum the taxon described matches R. ponticum ssp. baeticum in every consistent character; As Dr Richard Milne comments (personal correspondence) ‘there is no morphological trait that consistently distinguishes British R. ponticum from native Iberian material; instead we have occasional traits like hairy ovaries and long calyx lobes that are different, but as not all individuals have them, it cannot form the basis of a formal description.’
There is therefore simply no basis for accepting R. x superponticum as a valid taxon and I therefore propose that it be reduced in synonomy with R. ponticum ssp. baeticum or better still, just abandoned. As has been discussed, there is so far no evidence that any hybridity in R. ponticum populations make it more invasive. As Dr Richard Milne pointed out to me, even if such evidence were found, it would still not justify a new taxon name, because consistent distinguishing characters do not exist. The Cullen paper is responsible for a great deal of misinformation (you’ll find lots of articles which refer to it on the internet) and it was featured by James Wong on BBC television in The Great British Garden Revival in January 2015. I’m hopeful that this paper may go some way to redressing the balance and lets just put the ‘superponticum’ idea to bed once and for all.
Rhododendron ponticum, a mauve-flowered European native rhododendron, was introduced to the British Isles in the 18th century and planted widely as a colourful evergreen. It is now an out of control weed in western parts of the UK and is on the verge of being proscribed in both National and EC-wide Legislation so that it can no longer be sold or planted. I support this ban being applied in the UK as this plant is clearly out of control in some areas and in addition, R. ponticum is a vector for the spread of Phytophthora ramorum into forestry plantations.
In 2011 Dr James Cullen wrote a paper Hanburyana (Vol 5 June 2011) claiming that wild/naturalised Rhododendron ponticum in the UK is a hybridised superweed which he describes as R. x superponticum.
I read it at the time and found it unconvincing. I was also aware that Dr Richard Milne, based in Edinburgh, had already written a paper on this same subject, based on his study of the DNA of 260 samples of naturalised R. ponticum. James Cullen should have referred to the Milne paper in detail in his own paper as it covers with rigorous science what James Cullen does with outdated visual morphological study.
My belief is that Cullen’s paper and his ‘new’ species is based on unsound data, poor observation and minimal science and we should ignore its findings. Disappointingly the article has been widely quoted in the press and on line but not hitherto critically evaluated.
R. ponticum occurs in its wild habitats in isolated parts of Spain and Portugal and in larger populations in the mountains of E. Turkey, Georgia and adjoining countries. These very disjunct populations are probably the result of climate change following the last Ice Age. Fossil records suggest that R. ponticum was once widespread in Europe including western Ireland (Jessen, 1948) but that extreme cold and dry lowland conditions probably resulted in its extinction over most of Europe. It might well have (re-)established as a British native had the landbridge not been submerged creating the British Isles.
R. ponticum, one of 900 species of rhododendron,is the only rhododendron species or hybrid which is a major invasive weed in the west of the UK and parts of Brittany, France. Rhododendron luteum, the scented yellow azalea species has naturalised in some areas but is not causing anything like the environmental problems that R. ponticum has. Some west coast woodland gardens in the UK such as Stonefield in Argyll have self-seeding of many species and hybrid rhododendrons but these have not escaped the bounderies of the garden and naturalised in the countryside. Ledum palustre (now Rhododendron tomentosum) may be native to the UK and can be found in bogs in central Scotland while North American Ledum groenlandicum (now Rhododendron groenlandicum) is naturalised in parts of the peak district.
R. ponticum has the ability to spread by seed or layers and can regrow from roots or small parts of stems so it hard to eradicate with physical or chemical means. In areas with high rainfall and acidic soil it spreads rapidly and can form a virtual monoculture. In recent years it has become susceptible to Phythopthora ramorum and P. kernoviae, particularly in combination with Larix kaempferi plantations and this had led to greater efforts to eradicate it.
In October 2014 I went to study one of the best known wild populations on the mountain of Foia (902m), Monchique in the Algarve in Portugal. On the north side of the mountain there are substantial populations of R. ponticum growing in open hillsides with Cistus and Rubus. The population was flowering freely, producing plenty of seed and young seedlings were observed. In every way, morphologically this population matches the R. ponticum I have observed in many sites in the UK from Cornwall to North West Scotland.
Visible characters include:
1. Dark green narrow pointed leaves,
2. leaf retention for many years,
3. often red stems,
4. narrow curved seed capsules.
R. ponticum on Foia, Monchique, S Portugal, (902 m) October 2014 showing the typical narrow deep green leaves and red stems.
In 1962 my father Peter Cox and Peter Hutchison collected R. ponticum in Turkey. (AC&H 205) These collections have proved to be quite distinct morphologically from the R. ponticum from the Iberian peninsula and thus far have not shown any propensity to seed, layer or spread from where it was planted at Glendoick over 50 years ago. It is possible that if allowed to cross with Iberian forms, evolve further and be equally invasive but again there are no examples of this at Glendoick. Dr Richard Milne and Fatima Sales observed significant variation in the Turkish R. ponticum in their field studies and they concluded that we only have a small part of the wild gene pool of R. ponticum in cultivation.
R. ponticum AC&H from the Caucasus.
Milne & Abbott (2000) examined samples of DNA of naturalised R. ponticum and found that that all British and Irish naturalised R. ponticum is derived from Iberian stocks and not from Caucasus provenance. The Iberian form of R. ponticum has been described as R. ponticum ssp. baeticum to distinguish it from the Caucasus forms. The Edinburgh revision of Genus Rhododendron part authored by James Cullen did not recognise this subspecies. As the population in the Iberian peninsula is morphologically distinct and the disjunct populations are hundreds of miles apart, it would be sensible to maintain this subspecies. This should be the name used to describe the UK naturalised R. ponticum and not R. x superponticum. The Caucasus forms would be known as R. ponticum ssp. ponticum.
R. ponticum and several north American rhododendron species including R. catawbiense and R. maximum were introduced in the 18th century to the UK. Almost immediately, nurserymen began crossing the species and creating hybrids which were planted in gardens all over the UK.
James Cullen was an ‘old school’ taxonomist who did not use DNA as a tool to his work, despite its availability. Cullen therefore relies entirely on morphological characters (i.e. characters than can be seen). He lists errant morphological characters observed in some of the sample material gathered as R. ponticum for his study from many locations in the UK and suggests putative parent species:
1. Pedicels with dendroid hairs: 4 specimens. This is a character of R. catawbiense.
2. Calyx more than 3mm: 21 specimens. This is a possible character from R. maximum.
3. Ovary rounded, hairy (generally red or brown tomentose): 5 specimens This is a character of R. catawbiense, R. maximum (and R. macrophyllum.)
4. Ovary rounded, hairy and glandular: 2 specimens. This is a character of R. maximum.
5. Leaf length/breadth ratio less than 2.5, the apex and base rather rounded: 3 specimens. A character of R. catawbiense’
Dr Richard Milne, based at University of Edinburgh sampled the DNA of R. ponticum from samples taken from many sites in the British Isles in his 2000 study. His conclusions were:
1) There are genetic markers for R. catawbiense spread across many British R. ponticum populations at low frequency, and that these are more common in colder regions.
2) The hairy ovary found in some specimens of UK R. ponticum, suggests some R. catawbiense involvement.
3) R. maximum appears to transmit occasional long calyx lobes to some populations of naturalised R. ponticum but these are not common.
3) Red corolla flecks in some specimens suggest involvement of a species outside subsection Pontica. This is probably R. arboreum, introduced in the 18th century from India and crossed many times with R. ponticum to create swarms of hybrids known as ‘Altaclarense’, ‘Cornish Red’ and ‘Sir Robert Peel’ These hybrids are common in Argyll, Ireland and Cornwall in old gardens where they form the largest rhododendron bushes known in the UK.
So let’s accept therefore that some North American rhododendron species genes are involved in small quantities in some naturalised UK R. ponticum populations. The key here is to ascertain whether the R. maximum or R. catawbiense genes would make R. ponticum more or less invasive? We might be tempted to assume ‘hybrid vigour’ but there is no evidence for this.
My father Peter Cox and I have seen little sign of these hybrid morphological characters in the many naturalised R. ponticum swarms that we have examined in Scotland, Ireland and England but if you look hard enough you can find occasional probable hybrid traits such as hairs on the ovary, suggesting some North American genes as both Cullen and Milne have stated. But from what we have observed the most aggressively invasive R. ponticum seems to exactly match the Iberian wild populations. Most of them look like 100% R. ponticum with no other species involved. The relatively small number of hybrid plants we have observed in such populations tend to stand out and show no signs of being more aggressively invasive.
Pure Iberian R. ponticum appears to be perfectly hardy in the UK climate. R. ponticum regenerates freely from damaged stems, and roots and is therefore more than capable of withstanding the effects of most cold winters in the UK, particularly in the last 40 years where winters have been milder. I observed no more than superficial foliage damage to pure R. ponticum in the 2010-11 winters which saw record low temperatures for much of the British Isles for long periods. R. catawbiense is a very hardy species so it would add extra hardiness to R. ponticum. However R. catawbiense is poorly adapted to the UK climate and does not grow easily or well. The same applies to R. maximum. Both these North American species have evolved to withstand the hot summer and very cold winters of a continental climate in E. USA and in the UK have weak root systems, chlorotic foliage and little potential as garden plants. They would not confer any genetic or evolutionary advantage to a species which already ideally suited to the climate of the British Isles.
Dr Richard Milne adds weight to this conculsion in his findings: (personal correspondence). Milne believes naturalised R. ponticum shows some evidence of introgression, through hybridisation and repeated backcrossing to R. ponticum, from R. catawbiense. This could create a situation where some British R. ponticum populations have on average something like 5% R. catawbiense in their genetic makeup. Such plants would look very like pure R. ponticum but show the occasional morphological or molecular markers of R. catawbiense. Introgression is a well known process in many wild plant populations and is common in wild populations of rhododendron species. Dr Richard Milne points out that natural selection during the introgression process would probably remove genes with a negative effect on fitness, so that after generations of seeding and maturing, eventually the only R. catawbiense genes left behind would be those with no fitness effect, or those that might confer an advantage. He suggests that such genes could lead to enhanced cold tolerance, but that we have as yet no evidence for this. It is possible that R. ponticum may evolve some hybrid vigour via introgression, but as it seems so successful an invader already, there may be no evolutionary need to do so. Cullen makes the mistake of observing populations where some hybrid characters are apparent in some populations and then assuming that all R. ponticum naturalised in the UK should therefore be regarded as hybrids. There is simply no scientific evidence for this. It is at best a hunch. For which he provides little or no evidence.
Another angle is to look at garden-raised/man-made hybrids of R. ponticum to see if they show any invasive tendencies. There are numerous man-made hybrids containing R. ponticum crossed with North American species. Most of the hybrids in purple and lavender shades are this mixture. Cultivars such as 'Fastuosum Flore Pleno', ‘Blue Peter’, ‘Old Port’, ‘Lee’s Dark Purple’ and many more. None of these shows the slightest tendency to invasiveness by continuous layering or self seeding as R. ponticum does. This is not surprising, given that North American species are not well suited to cultivation in the UK, tending to chlorosis and lack of vigour. My observation is that far from crossing with R. ponticum to make it more vigorous and invasive, the reverse is true: adding genes from these species seems to bring R. ponticum under control. In other words hybridisation into R. ponticum of any of these species makes it less invasive rather than more invasive. So the epithet R. x superponticum is therefore a misnomer.
Cullen, Dr James ‘Naturalised rhododendrons widespread in Great Britain and Ireland’ Hanburyana (Vol 5 June 2011)
Jessen K (1948) Rhododendron ponticum in the Irish interglacial flora . Irish Naturalists' Journal, 9, 174 175.
Milne, R. I. & Abbott, R. J. Origin and evolution of invasive naturalised material of Rhododendron ponticum L. in the British Isles May 2000: Molecular Ecology. 9, p. 541-556
Kenneth Cox is one of the world's leading rhododendron experts and author of 7 books on Rhododendrons including the monograph The Encyclopedia of Rhododendron Species.
Ken runs Glendoick Gardens and Garden Centre and is Scotland's best selling garden writer.
