Comment on 'BATCH STUDY of C. miniata × C. nobilis CROSSES'.
[Clivia News 20 (1): 11-12 (Jan-Mar 2011).]
I get seed on my clivias all the time, without doing anything. I have never seen the stigmas of my plants of Clivia miniata in direct contact with any anthers, nor have I seen any pollinators ever visit my flowers. My umbels produce berries, except usually on the outermost flowers, with those berries that contain the most seed being plumb in the middle of the umbels. The pollen is thus crossing the gap between the anthers and the stigmas on its own two legs.
I believe that the C. miniata plants in my greenhouse are anemophilous (which is not a Latinised version of the name of that astounding Italian-American coloratura soprano, Anna Moffo, but rather means 'wind pollinated' - anemoV (anemos) = wind Gr. ; filia (philia) = love Gr.). Wind pollination is a very successful strategy to adopt - all grasses do it, so do palms, conifers and casuarinas, amongst many others; and their specially-designed pollen can fly for miles.
Perhaps C. miniata is also anemophilous in the wild? While it is quite easy to visualise that the tubular, pendulous species of Clivia, with their gathered stamens and stigma, are perfectly designed for pollen to be collected on and subsequently deposited from the faces of visiting sunbirds (this has been well covered by John Manning in Clivia 7: 17-25), I have no idea what animal could possibly be involved in the movement of pollen between the upward-facing flowers of C. miniata. Those stiff, upright filaments carry their wobbly anthers like aerials, with the stigma protruding out of the middle of the flower, although it can be decumbent (reclining or lying) in horizontally directed flowers.
In other infundibuliform (funnel-shaped) amaryllid flowers, like hippeastrums and Amaryllis belladonna, the stamens are also usually decumbent, presenting themselves in a bunch at the lowest point of the flower, which makes pollen contact with insects a much easier proposition. The horizontally directed flowers of some C. miniata clones do, in fact, present with decumbent stamens - particularly the peaches and green-centres. This probably indicates that these groups are introgressed with genes from pendulous clivias, whose filaments are less rigid, as they are only required to hang downwards and not perform any mechanical feats.
I cannot imagine anything that can collect the pollen from the stamens of an upright-flowered C. miniata and deposit it on a stigma. No birds are likely to do it, as perched on the scape, they cannot easily reach the flowers. Bees may visit flowers and wonder all over them, but they are too small to reach the tips of the anthers without climbing them, except some giant carpenter bees (as well as the butterflies proposed by Manning) - and then where would the pollen be placed upon the insect in such a way that it could reliably land up being deposited on a stigma? The architecture of this flower does not offer the facility required for the precise placement of pollen - something I believe is fairly important in most animal-pollinated flowers.
But, having stigmas pointing skyward, these flowers offer the perfect landing pad for airborne pollen grains. As the stigmas of the pendulous species point downwards, wind pollination would be rather difficult for them to achieve. The horizontally-directed flowers of C. miniata also present less of a pollen-receptive surface for air-borne pollen, for purely mechanical reasons, and I assume that this would be the reason why I get smaller or no berries on those flowers furthest from the centre of the umbel. I thus theorise that the space around flowering C. miniata plants comprises a nebulous cloud of predator pollen, waiting to sexually attack innocent virgin stigmas.
Of course, the only thing in biology that is cut and dried, is sliced biltong. The pendulous clivias are scentless, but there are clones and strains of C. miniata that have a fine azalea scent. Plants don't do things for no reason, or simply to impress humans (they could care less about people!), so that scent must be there to attract an insect (birds do not respond to floral scents). Until I see the evidence that some large insect does the job properly, I will have the temerity to suggest that honey bees are attracted to such fragrant flowers in order that the egg-beater action of their beating wings can be employed by the plant to liberate its pollen into the aerosol cloud! (What could be more radical or bizarre in the field of pollination biology than to suggest the existence of 'bee-mediated anemophily'? - You heard it here first!)
Based on the above, I believe that it is extremely difficult for a clivia breeder to create pedigreed plants without taking extraordinary precautions, such as: isolating breeding plant; emasculating all flowers before the anthers have had a chance to dehisce; and keeping the doors and windows closed. Anything less than that may very well lead to promiscuity. No wonder, I muse, that 'normal Mendelian genetics seldom works' in C. miniata. I am firmly of the opinion that we have far less control over the hand pollination of C. miniata than we fool ourselves into believing, unless we follow the gold-standard protocol of Willie le Roux (Clivia News 18 (3&4): 30), with his innovative 'clivia female condoms'; or we only own one clivia.
One illustrative example of where 'normal Mendelian genetics' appeared not to work, was with the late Cynthia Giddy's 'Natal Yellow'. Despite the fact that she selfed this clone regularly and raised some 500 seedlings, not a single yellow-flowered plant appeared among the progeny. She was not aware that this clone was self-incompatible and assumed that the collected seeds were the product of the labours of her pollination efforts - something obviously precluded in self-incompatible plants. The copious seed production was in all likelihood the product of anemophily from neighbouring orange plants, because correctly bred, this plant does yield yellow-flowered progeny.
A second example is that of the establishment of the New Dawn strain by Jim Holmes, as recorded by Harold Koopowitz in his book Clivias, p. 239. Referring to the original yellow × yellow cross upon which this strain was based, Koopowitz says: "Equal quantities of yellow and orange flowers were obtained in the first generation, and when the yellow offspring were intercrossed the numbers of orange-flowered plants dropped to between 5 and 10 percent. After a few generations, the offspring are all now yellow." My belief is that group1 yellow × group1 yellow = yellow. The percentage of orange progeny is a function of the size of the population of orange-flowered clones sharing the growing space with the yellows that are being bred. As I interpret it, the Holmes collection of yellows grew, his growing space became progressively less populated with oranges and the amount of available air-borne 'orange' pollen dropped, until eventually he was growing all his yellows together in an exclusively yellow house - and hence no orange-flowered progeny. The appearance here of orange plants had nothing to do with yellow × yellow genetics and everything to do with environment. The early history of yellow clivias is peppered with reports saying that: 'when selfed, the seedlings are orange'.
Having here once twisted the sobriety of conventional pollination biology, I suddenly feel the urge to do it once more. Perhaps the horizontally-directed flowers of C. miniata are more easily and effectively visited by insects, being more conventional in presentation; and pollination is then accomplished. Then my next model suggests that the upward facing flowers are mostly wind-pollinated - frequently selfed in self-compatible plants; and the edge-flowers are insect-pollinated and perhaps mostly outbred. This would give us the best of the two worlds - some inbreeding and some outbreeding - and a plant utilising two very different breeding systems in the same inflorescence! (You also heard that here first!)
Before I finish my set up, we must discuss the subject of self-sterility - more accurately called sexual self-incompatibility. It seems likely that the pendulous species all show some degree of self-incompatibility, requiring pollen from another plant of a different clone in order to set seed properly. Self-incompatibility is common in plants that produce scapes of many flowers opening at one time, all of which may be visited by a single pollinator - nectar-seeking in this case. It is a mechanism for the plant to reject or retard the growth of deposited own pollen, but allow for the germination of pollen from a different, earlier-visited scape of another plant, thereby obviating self-pollination. If the pendulous species did not exercise some self-incompatibility, they would simply self themselves in genetic oblivion, because a single visit by a sunbird is likely to result in own pollen being smeared onto the flower's stigma in every case. Self-incompatibility does not have to be an all-or-nothing mechanism. Simply slowing down autogenous pollen germination and growth, thereby allowing for a foreign pollen grain to win a race, is sufficient for a successful outcome. Thus selfing of such 'self-incompatible' clones may be possible, although it is likely that the selfed pollen may take considerably longer to effect fertilisation than would be the case were foreign pollen of the same species to be involved. In the time-scale of the clivia breeding cycle, such a difference would be negligible and thus not easily detected.
C. miniata is variable in this regard, with some lines like the Group 2 yellows being self-incompatible, but also having many lines which are self-compatible to some degree. It is likely that the earliest upward-facing flowers of the C. miniata evolutionary line had to be somewhat self-compatible in order to procreate at all, there being few 'original' specimens of this group, or perhaps even only one. Even self-compatible C. miniata plants are likely to exercise some aspects of self-incompatibility, as this is integral to the genus as a whole. Perhaps the mutation that allows self-compatibility of the haploid pollen grains of C. miniata, when present in the homozygous state in the diploid stigma (presumably a relatively common condition), creates a compatibility barrier against the pollen of some other Clivia species. This may also mean that only the styles of partly of fully self-incompatible clones of C. miniata, which are therefore not homozygous for self-compatibility, could allow for the growth of pollen from these other species.
Certainly such a genetic mechanism could conceivably have been part of the necessary evolutionary development of a taxon so radically different from its congeners, in order that it could develop and maintain its own species integrity. The biggest threat to such species integrity would perhaps have been the air-borne pollen of the other species with which it shared an environment. Such an acquisition by C. miniata would make complete evolutionary sense.
Unilateral interspecific incompatibility is a recorded phenomenon that occurs when pollen of one species is rejected on pistils of a related species, while no rejection occurs in the reciprocal cross. Although the successful crosses usually occur only when the self-compatible parent is used as a female, there are a very few exceptions in which self-incompatible pollen will not grow in self-compatible styles, the latter being the case when self-compatibility has originated only recently from self-incompatible species, something we can possibly assume is the case in C. miniata. This form of unilateral interspecific incompatibility would reinforce species identity in natural populations of related taxa which share the same habitat.
So after that long preamble, I finally get to comment on the results of Welland Cowley's C. miniata × C. nobilis crosses. The scenario I have laid out above allows for an explanation of what may very well be going on here. Many of Cowley's C. miniata pod parents exhibit unilateral interspecific incompatibility to C. nobilis (being homozygous for self-compatibility), and in the absence of a valid pollination or having pollen growing very slowly down the style, have accepted anemophilous 'predator pollen' supplied by other C. miniata plants in the general environment. The small percentage of true hybrids have arisen notably from a Group 2 yellow of the self-incompatible type and from green-throats, that have, in all likelihood, inherited their colouration and possibly concomitant self-incompatibility from a pendulous species way back in their ancestry, and thus do not exhibit the interspecific incompatibility.
I find myself unable to agree with Prof. Spies that genomic imprinting could be so massive as to completely expunge the effects of a full genome of other genes. I am certain that gene silencing through epigenetic processes is common and widespread in clivias - it must be - because the genes which create and organise the leaves are surely not those that create and organise the flowers, or the roots. However, I have a problem believing that the wide range of genes situated on an array of the chromosomes that have been derived from C. nobilis could ever be silenced by genomic imprinting in C. miniata. Whereas it is known that the triploid endosperm, so massive a part of the clivia seed and one which would be derived in these cases from two genomes donated by C. miniata and one from C. nobilis, is intimately involved in genomic imprinting, I believe that its effect could never be as great as this. That would imply that the genes of one species are labelled in some way differently to to that of the other species, in order that the cell would know which genes to express, and which not to. Cowley lists among his statistics: "Number of plants found with C. miniata vegetative or floral characteristics: 100". Cowley clarifies this further (pers. com.) by saying: "The assessment was made on plants which had not yet flowered - vegetative characteristics were examined and plants that had already flowered showing both vegetative as well as floral characteristics. As not all the plants were ready to flower, this is an early assessment, but so far it seems safe to say that if a plant displayed miniata-like vegetative characteristics it is likely that the plant will have miniata-like flowers." This indicates to me that these progeny show no characteristics at all of C. nobilis.
Cowley also comments regarding the opposite cross, C. nobilis × C. miniata, that "the progeny are usually plants with interspecific characteristics"; and so they ought to be, as C. nobilis is not apparently affected by any interspecific incompatibility or anemophily.
As far as I can see, selfing of 'this first generation of C. miniata-like plants' will give a complete generation of more C. miniata-like plants - which like the first generation, will in fact simply be C. miniata (in a broader sense, but without any F1 input of C. nobilis) - nothing more, nothing less.
My house of cards is set up upon one supposition after another supposition, and of course I would love to have you bash it down to the ground - that is why I present it to you. Whereas I may be an abstruse, annoying, argumentative, arrogant ass, I think that science becomes its most interesting, when results do not confirm one's opinions (because then you have to find a new way to explain them, logically). I have certainly considered the work that will be required in the coming flowering season to demonstrate or allow to be rejected some of the airy-fairy ideas I have suggested above, and I hope I will succeed in accomplishing it.
While some clones of C. miniata may exhibit unilateral interspecific incompatibility with C. nobilis, perhaps this mechanism is less active in those pendulous species more closely related to C. miniata; with C. gardenii seemingly showing little interspecific incompatibility.
What is most obvious to me from all the work I have done to construct this story, is that we know absolutely nothing about the breeding systems operative in the species of the genus Clivia. There is some random empirical evidence that various degrees of sexual self-incompatibly exist, but the limits are less than nebulous. Much study is therefore needed before we actually garner a guaranteed understanding of what is going on in such cases as that offered by the fascinating results that Cowley has encountered in his C. miniata x C. nobilis crosses. ::
Acknowledgements: Thanks to Welland Cowley for some clarifying correspondence, and I apologise to him that I have not simply discarded this entire article - something he thinks would be a really good idea. Thanks also to Roger Fisher for looking over this piece. From his response, I think that I should not push him to publish this on paper, as it truly does contain too much supposition and I may leave some one else with egg on their faces - when all the egg splashes should be directed at me. Prof. Johan Spies also looked over this piece, but being the scientist he is, could barely be drawn on commenting on such theoretical, but unproven, work.
© 2011 Greig Russell
If you don't like what I have written here and would like to give me hell -