Logo for the Journal American Rhododendron Society

Journal American Rhododendron Society

Current Editor:
Dr. Glen Jamieson ars.editor@gmail.com

Volume 58, Number 1
Winter 2004

DLA Ejournal Home | JARS Home | Table of Contents for this issue | Search JARS and other ejournals

Incompatibility in Rhododendron pachysanthum
Joe Harvey
Victoria, British Columbia

        Every year I try to send some interesting seeds to the Seed Exchange. My contributions comprise mainly F1 hybrids between indumented species aiming for foliage interest. Among the indumented species I use, my specimen of Rhododendron pachysanthum flowers well and is a source of abundant seed when pollinated with pollen from other species. Contrasting with this abundance of hybrid seed is its failure to produce seed when selfed despite attempts over a number of years.
        Now failure to set seed on a truss is not usually worth mentioning; one assumes something went wrong be it inopportune cold, rain or other accidents of nature and tries again the next year. But Rhododendron pachysanthum makes a handsome specimen itself and seed of it is in demand so I was sorry not be able to be contributing any.
        After three years of producing no pachysanthum seed I came to suspect that it wasn't just bad luck or incompetence, that there must be a reason behind the problem. It struck me that what I had was a case of self-incompatibility. This is fairly common in hermaphrodite plants where a plant cannot set seed when pollinated with its own pollen but readily does so when pollinated from a flower of another individual of the same species. "Individual" in this case means a plant produced from a separate seed - clones produced by means of grafts or cuttings all count as one genetic individual. The phenomenon is best known to gardeners in the need to plant different clones of sweet cherries or pears to produce fruit. However, there is little known about the situation in rhododendrons.
        Being a retired scientist - and scientists never retire - I concocted a simple test to detect self-incompatibility. I took two trusses on my plant, selfed the flowers on one truss and cross-pollinated those on the other, all the while taking precautions by means of stigma caps to prevent any unwanted pollen from reaching any of the flowers.
        My specimen of Rhododendron pachysanthum was obtained from the Rhododendron Species Foundation as 78/064 but the problem was to find a genetically distinct one. Fortunately I mentioned my difficulty to colleague Norm Todd who happened to have a plant in flower. Although the source of his plant had been lost it was obvious that it was distinct since the indumentum on the upper surface of the young leaves was silvery, and on mine golden. The indumentum colour is constant from year to year and appears to be genetically controlled although heavy overhead irrigation or rain can reduce or remove it. I therefore used the Todd plant as the source of pollen for the outcrossed truss.

Truss No. of capsules Grams of seeds Grams of seeds per capsule No. of seeds per seeds per capsule
Selfed 10 0.06 0.006 40
Outcrossed 15  1.38  0.092 613

        Initial development of the capsules was disappointingly similar in both trusses; they all began to swell. However, by midsummer those of the selfed truss stopped growing, but the others continued to enlarge and were conspicuously plump by the time they ripened in mid November. To compare productivity I extracted and cleaned the seed and weighed it. The results are plain to see: seed production from the outcrossed flowers is fifteen times greater than from the selfed one. I determined the average weight of a seed to be 0.00015g by counting out 200 seeds and weighing them.

R. pachysanthum seed pods
R. pachysanthum seed pods.
Photo by Joe Harvey
R. pachysanthum, underside 
of new and old leaves.
R. pachysanthum, underside of new and old leaves.
Photo by Joe Harvey

First, the case for self-incompatibility seems proven. The fact that the incompatibility was not absolute, that a few seeds were formed on selfing, corresponds to observations in other plant genera where the particular genetic system involved can allow some "leakage."
        The ecological function of self incompatibility is that it is an outbreeding mechanism that serves to maximize genetic diversity in the population. In the wild where only a few species will be in flower at any one time and where each may have a different spectrum of pollinating agents this is a reasonable way to maintain the gene pool in a stable population. However, it does mean that the occasional wayward insect may transfer the "wrong" pollen to a stigma and natural hybrid seed will result. This has been observed occasionally in seedling batches grown from wild collected seeds and the usual vigour of these hybrid seedlings encourages the grower, often unconsciously, to select them for growing on. The confusion that results can last a century.
        In nature hybrids usually fit badly into the habitat and do not survive to reproduce except in cases where the habitat has been disturbed by fire, logging, grazing, etc., and Edgar Anderson's adage that "hybrid plants fit into hybrid habitats" applies. By hybrid habitat he meant disturbed sites where growing conditions differ from adjacent more natural areas. This is the reason for the occurrence of rhododendron hybrids on areas which have been logged or are abandoned farmland around the "balds" in the Appalachians.
        For people who raise rhododendron species from seed there are implications involving their seed sources. Wild collected seed as mentioned has a low but measurable chance of producing hybrids rather than the species from which they were collected. The picture changes when one attempts to grow species from capsules on open pollinated plants in gardens, particularly botanic gardens where a great diversity of species has been brought together in an unnatural situation. Seed from those plants which are self-incompatible are almost certain to yield some or all hybrids. We don't know which these plants are but the phenomenon is by no means confined to Rhododendron pachysanthyum. On Taiwan three species of Rhododendron in Subsection Maculifera occur on the mountains apparently differing somewhat in their altitudinal and habitat distribution. In addition to Rhododendron pachysanthum of the current discussion they comprise R. pseudochrysanthum and R. morii. My observations on R. pseudochrysanthum in gardens is that it sets seed abundantly and comes true and it would seem to be self-compatible. On R. morii I have no observations. However, it is noted in Cox and Cox, The Encyclopedia of Rhododendron Species, that hybrids have been found in the wild between all three species and this is to be expected on an island under fairly intensive human exploitation.
        From a wider perspective, incompatibility systems play an important role in plant spread and have been particularly studied on islands, especially the Hawaiian Islands. When a plant is spreading into an area, say terrain left after an Ice Age, self-incompatibility is a disadvantage to rapid colonization since a seed, which by a fluke manages to get distributed a great distance from its parental population, cannot set up a new colony because of its inability to produce much or any seed. Under these circumstances self compatibility - the ability to produce abundant seed from a single plant - is at a premium. This is demonstrated by Rhododendron macrophyllum in western North America which was able to spread from what has been shown to be two survival areas after the most recent Ice Age. Rhododendron macrophyllum is an extremely seed-prolific plant even from a single specimen.
        The situation on islands is similar but different. New islands, such as volcanic islands which rise out of the ocean, receive seeds by rare, long distance transport events and in these circumstances usually only one seed makes it and becomes established. Thus initially the plant has to be self-fertile if it is to reproduce and set up a population. At this point another genetic force comes into play and self-incompatibility becomes useful to ensure outcrossing and hence maximizing the gene mixture in the offspring. This has been demonstrated in several genera on the Hawaiian Islands but not alas in Rhododendron.
        Taiwan is not such an isolated land mass as the Hawaiian Islands, but it seems fair to speculate that three separate events, at different times, led to the introduction of Subsection Maculifera propagules from mainland China and that these evolved over time into the current three species under the influence of the different soils, climate, insects, etc.
        The self-fertile Rhododendron pseudochrysanthum may be the surviving remnant of a formerly widespread species on the Chinese mainland colonizing Taiwan during a cooler climatic episode when the sea level was lower and the Taiwan Straits narrower. In this sense it may parallel the southern survival R. macrophyllum but without the ability to move back to the mainland after the Ice Age. The Ice Age affected southern China less than North America and there was little unoccupied habitat to recolonise after it.
        Rhododendron pachysanthum seems a better candidate for the classic long-distance transport by one or very few seeds and the evolution of its self incompatibility in situ on Taiwan. This is speculation and we need to know what are the closest relatives on the mainland from which the three species might have evolved.

Joe Harvey is a member of the Victoria Chapter.

Volume 58, Number 1
Winter 2004

DLA Ejournal Home | JARS Home | Table of Contents for this issue | Search JARS and other ejournals