With few exceptions, and those principally in the Tillandsias, the flowers of the bromeliads last but a few hours. Other exceptions to this rule may be found in many of the Vriesias. If they blossom in cool weather the flower may carry over to the second day.
Some species of the bromeliads flower early in the morning and may be gone by midday. One species of Aechmea which I found in Brazil baffled me for several days. I finally learned that the petals opened at midnight and by three A. M. those petals were completely disintegrated. By way of contrast the flowers of Tillandsia crocata last for four to five days according to the prevailing temperature.
When the pistil and stamens are very short and are far down in a slender corolla of the flower they cannot be seen without carefully removing the sepals and petals so as to expose these reproductory organs. This is the condition in the flowers of Tillandsia lindenii or T. cyanea.
Some of the giant Hohenbergia species such as H. Saltzmanii or H. Blanchetii, have very small flowers; the petals may be less than an eighth of an inch long and the flower may be completely filled with nectar. All of these different conditions may be met with if you are interested in the pollination of the bromeliads, but be assured that if you are interested enough to study these different conditions in bromeliads and plant life in general will carry a great and lasting compensation to the better understanding of all life in its myriad manifestations; you will find parallels everywhere.
Once you have determined the time of day that the flower is completely open and the stamens and pistil are in mature form, then you may procure the pollen on a small camel's-hair brush and carry it to the flower of the plant which you may desire to fertilize.
Your problems of pollinating the flowers will vary with different species. In many flowers you will find the stigma of the pistil extending out beyond the stamens. Such a flower is very easy to pollinate, but in many species the stamens may exceed the pistil or may be of even length. In many flowers the stamens may completely surround the stigma so that they would have to be removed in order to pollinate it with the pollen from another flower. If this be the case, then Plant No. 1, the maternal plant, should be emasculated as soon as the flower is open or mature. This means that the stamens must all be removed very care fully so that none of the pollen falls on the stigma to cause self-pollination. The pollen should then be carried to it from Plant No. 2, the paternal plant. This may necessitate a little practice with a steady hand. If you wish to use Plant No. 2 for the same cross, then the same care should be used in removing those stamens. Now the pollen is carried from Plant No. 1 to No. 2 which will make the same cross with the same results, as each flower has the parts of both sexes.
Should the two Species which you wish to cross, mature their flowers at different times of the day, you may find a short period between those two maturity hours that you can pollinate one stigma before it is entirely receptive and the other before it has entirely passed that period and still catch both flowers. But if the time space between the two maturity periods is too great, you may be able to keep the pollen from one plant by taking the stamens and keeping them in waxed paper or in a tightly sealed bottle in the refrigerator for a few hours. I have not been able to keep the pollen of bromeliads for any great length of time, but further experiments may make this a possibility. With some flowers in other families, pollen has been kept for a period from six months to a year.
Should your two subjects be two different true species, then the results of this hybridization will be rather constant. Variation in the offspring will not be very great. However, if one of the parent plants in this cross pollination should be a species and the other a hybrid, or if they both be hybrids, then the offspring will run a most interesting lot of hybrids with the different combinations of character of parents and grandparents.
If you wish to self-pollinate a plant for the propagation of more of the same species, then you carry your brush from flower to flower back and forth from plant to plant, thrusting carefully your brush into each flower, being sure that each stigma is covered with pollen.
Rarely will your bromeliad flowers under cultivation be pollinated unless you do it artificially, as there appear to be few insects that are attracted to them. We believe that most of the pollination of bromeliad flowers in their natural state is performed by humming birds and night moths. In some instances, how ever, the ants and bees have been the pollinators, especially with flowers that carry considerable nectar. The pollen is not dispersed by wind because it is of a semi-moist nature.
Then the rare exception of self-pollination may be found in such species as Guzmania graminifolia and other new species of Guzmania which I found in Colombia in 1946. The flower heads on these plants were entirely encased in a jelly-like substance, thus preventing any insect from entering the flower. Also, there are a number of other species that bear flowers which never open, and yet they produce seeds without any outer agent of pollination. In these instances, no doubt, the flowers are sell-pollinated.
Most species of bromeliads have a number of flowers on each inflorescence, but the range in quantity may vary from a single flower, as in T. usneoides to as many as 8,000 in Puya raimoudii This giant Puya has a flower head twenty feet long and eight feet in circumference which holds these myriad flowers.
Nearly all the bromeliad flowers function as perfect flowers, that is, they contain the reproductory organs of both sexes. However, there are exceptions. The Hechtias have both parts, pistil and stamens, but only one sex can function in each flower. Some plants, considered as male plants, will have flowers in which the stamens are perfectly formed while the pistil is aborted and incapable of receiving pollen. In the female plant the flowers have aborted, undeveloped stamens but have a pistil fully developed and capable of accepting the pollen from another plant. A similar condition may be found with some of the Catopsis and Dyckia species.
With the Cryptanthus species there is still another condition. With very few exceptions, and I have examined hundreds of flowers, I have found only male flowers in the center cluster. These flowers are imperfect and have no trace of female parts, but the flowers in the axils of the bracts below the center cluster are invariably perfect, having both sexes that function. The pollen from the center male flowers is fertile and can be used for pollinating, but these flowers are generally spent before the lower perfect flowers appear. The lower flowers can be self-pollinated or can be used for cross pollination in hybridizing.
Self-pollination by hand is not always successful with some species, as it seems that some of them are self-sterile. This condition is still more prevalent among the hybrids. However, it is often the case that, while self-pollination among the flowers on a single infloresence may not be successful; the pollen from one of those flowers may possibly be successfully used on another similar plant, species or hybrid.
Not always are the results of hybridization an improvement over the parents used, but in most instances they will give the hybridizer an interesting new combination of characters quite worth while. I doubt if anyone can foretell which parent plant will influence the resultant hybrid most. However, after attempting several different crosses by using Aechmea Weiibachii as one of the parents, I have as yet to find even one of those crosses to be of any value.
Bigeneric crosses can be most interesting although but few such crosses have survived in horticulture. Some well worth-while Cryptbergias have been developed. The cross between Cryptanthus beuckerii and Biilbergia nutans has become common. As in many bigeneric crosses, the flowers are aborted and have little attraction, but as a foliage plant it has a definite decorative value.
Until 1930 little had been done in hybridization among the bromeliads excepting in the genera of Cryptanthus, Billbergia and Vriesea, and most of that hybridization was carried on in Europe. Even up until the last ten years, little had been done outside of Europe along this line except in Billbergia and Cryptanthus by Florida and California hybridizers.
Now with all the new species that have been introduced into horticulture in the past twelve years, the possibilities have been greatly enlarged with a most interesting field at hand for horticulturists to enjoy the privilege and experience of assisting in the creation of new, decorative forms and color combinations in the Bromeliaceae.