Hallé, F., 1978. Architectural variation at the specific level. In: Tomlinson, P. B., & Zimmerman, M. (Eds.) Tropical Trees as Living Systems, 1978. pp. 209-221 Architectural variation at the specific level in tropical trees Francis Hallé, 1978. Institut de Botanique, Montpellier, France My main research is a genetic approach to plant form in its entirety aimed at improving our understanding of the inheritance of plant architecture. It has been established that vegetative architecture is a constant and stable characteristic of the plant (Hallé, 1971, 1974; Hallé & Oldeman, 1975). Of course, ecologic conditions often involve some changes of form, but these are only quantitative, the basic architecture remaining unaltered. For instance, the commonly cultivated tree Muntingia calabura L. (Tiliaceae, tropical America) has quite a different appearance depending upon whether it is growing in forest shade or in open secondary vegetation. Although the physiognomy shows a contrast between a tall, narrow-crowned (Fig. 8.1A) and a low, broad-crowned (Fig. 8.1B) individual, the architectural model itself is free from ecologic variations, at least within certain limits (see Chap. 9). As a general rule, therefore, we may consider the architecture as a constant character, valuable for specific identification; indeed, the architectural model of a species has been quoted in several recent identifications (Hallé, 1970, 1973; Mabberley, 1973, 1974, 1975). It is obviously one of our aims in this symposium to bridge the gap between taxonomy, the study of plant parts, and developmental morphology of the whole plant. Nevertheless, this general rule of architectural constancy at the specific level makes it difficult to explain the polymorphism that occurs in most families and even in some genera. For instance, as far as I know, one can find 8 models in the Moraceae, 10 in the Sterculiaceae, and 10 within the single genus Euphorbia L. (Cremers, 1975). The existence of this polymorphism in the higher taxa inevitably implies a mechanism of architectural variation at the specific level, and this gives rise to the question of the nature of such a mechanism. Of course, the stability of the architecture as a hereditary character is compatible" with a certain degree of variation; this kind of situation is well known in genetics. If tropical trees are living systems, and I am pretty sure they are, their architecture must permit some variability. Fig. 8.1. Two different forms of Muntingia calabura determined by ecologic circumstances, architecture remaining unchanged. A. Tree (arrow) in dense forest. B. Open-grown tree. Architectural modification. Observation demonstrates that a specific qualitative variation of the vegetative architecture does exist and, as a first step to.ward understanding it, I present some examples, divided into three groups: intraspecific variation correlated with sex, ecologically imposed variation, and variation due to mutation. Intraspecific variation correlated with sex. In Cycas circinalis L. (Cycadaceae, Southeast Asia) the in.traspecific variation is correlated with sex; male and female plants belong to different architectural models, as essentially shown by Chamberlain (1911). The female Cycas tree conforms to Corner's, the male to Chamberlain's model (Fig. 8.2). In the former, the trunk is a monopodium, with ovulate sporophylls borne laterally; in the latter, the trunk is a sympodium, each module ending in a male cone. Fig. 8.2. Architectural variation correlated with sex in Cycas circinalis. A. Female plant (Corner's model). B. Male plant (Chamberlain's model). Ecologically imposed variation. In several trees recently recorded by Kahn (1975) - Euphorbia mellifera Ait. (Euphorbiaceae, West Africa); Arbutus unedo L. (Ericaceae, Mediterranean region), Mangifera indica L. (Anacardiaceae, Southeast Asia), Isertia coccinea (Aubl.) Gmel. (Rubiaceae, Guianas) — the vegetative architecture is variable and obviously correlated with the environment, most closely with the amount of incident light. In full sun, these trees conform to Leeuwenberg's model; in the shade, to Scarrone's model (Fig. 8.3). The difference is established by the continued activity of the apical meristem to form a distinct trunk in the latter, and by its early abortion in the former. Fig. 8.3. Architecture influenced by environmental conditions. A. Open-grown tree, corresponding to Leeuwenberg's model. B. Tree of same species growing in shade, corresponding to Scarrone's model. For further explanation, see text. Under these circumstances we have a contrast between fully and partially modular construction (see Chap. 9). The amplitude of this variation is narrow, and the only conclusion we can draw is that there is a close developmental relation between these two patterns. Variation due to mutation. Much more interesting is the third group, where are gathered what 1 propose to call "architectural mutations." The term "mutation" seems to be accurate because there is a sudden appearance of the modified architecture, which, at least in some cases, is heritable. Most of the architectural mutations described below concern tropical crop trees. Observations on cultivated plants are important, as they enable us to study wide monospecific and homogeneous plant populations, in which mutants are easily detected. The following examples illustrate the concept of architectural mutation (Figs. 8.4, 8.5). Leeuwenberg's model is the normal architecture of cassava (Mani.hot esculenta Crantz; Euphorbiaceae, South America), as shown in Fig. 8.4A. The mutation "pétiolule" belonging to Chamberlain's model (Fig. 8.4B) was discovered by Cours (1951). The normal form of the travelers' tree (Ravenala madagascariensis Gmel.; Strelitziaceae, Madagascar) is Tomlinson's model, although it is often seen as Corner's model owing to artificial removal of the basal suckers (Fig. 8.4C). The mutation "apical sexuality," corresponding to Holttum's model, was found in Lae, Papua New Guinea, in 1972 (Fig. 8.4D). The opposite situation is also known, where Holttum's model is the normal form, as in maize and in tobacco (Fig. 8.4D). Several mutations (indeterminate growth) with lateral sexuality have been described in both plants: by Singleton (1946) in maize and by Jones (1921) in tobacco (Fig. 8.4C). Rauh's model is the normal architecture of Pinus and of rubber (Hevea brasiliensis Miill.-Arg.; Euphorbiaceae, South America), as shown in Fig. 8.4E. Kozlowski & Greathouse (1970) described the mutation "foxtail" in Pinus caribaea Maorelet (Pinaceae, Central America), without branching, without annual rings in the wood, but fertile. Male cones have been recorded on trees growing in Kepong, Malaysia (Fig. 8.4F). The same mutation, but sterile, appears in rubber, and is known as "lampbrush." It has been demonstrated by Hallé & Martin (1968) that the lampbrush mutation can be produced experimentally by removing most of the foliar surface when the leaves are young. Both foxtail and lampbrush belong to Corner's model. Fig. 8.4. Architectural mutations. A. Manihot escalenta, normal form, conforming to Leeuwenberg's model. B. Pétiolule mutation conforming to Chamberlain's model. C. Ravenala madagascariensis, shown here without basal suckers and conforming to Corner's model. D. Mutant form with apparent terminal inflorescence, ob.served in Botanic Garden, Lae, Papua New Guinea. E. Rauh's model, illustrative of the architecture of many trees. F. Lampbrush and foxtail mutations of Hevea and Pinus spp., a modification of Rauh's model characterized by lack of branching. In Pinus pinaster Ait. subsp. maritima H. del Vill. (Pinaceae, Mediterranean region; Rauh's model), a mutation called "piboteau," with apical female sexuality, was recently described by Dupuy & Guédès (1975). This Pinus piboteau belongs to Koriba's model (Fig. 8.5A). Fig. 8.5. Architectural mutations. A. Piboteau mutation in Pinus pinaster, which because of its sympodial trunk growth represents Koriba's model rather than original Rauh's model (cf. Fig. 8.4E). B. Roux's model, normal architecture of coffee. C. Plagiotropic muta.tion of coffee that corresponds to Troll's model. D. Petit's model, found in certain Malvales. E. Little- or unbranched mutation con.forming to Comer's model. F. Abortive terminal mutation of cot.ton. G. Similar mutation in Abroma augusta. Roux's model is the normal architecture of coffee (Coffea arabica L.; Rubiaceae, Ethiopia) as well as of Paliurus australis Gaertn. (Rhamnaceae, Mediterranean region; Fig. 8.5B). The plagiotropic mutation, corresponding to Troll's model, was discovered by Carvalho & An- tunes Filho (1952) in coffee grown from seeds (Fig. 8.5C). It is well known that a plagiotropic coffee is easily obtained by propagation of branches of the coffee tree. In Paliurus australis Gaertn., a sterile plagiotropic mutation has been experimentally produced by cultivation of etiolated plants by Roux (1968). The model of Petit (Fig. 8.5D) is normal for several members of the Málvales, such as jute (Corchorus capsularis L., Tiliaceae, pan- tropical), cotton (e.g., Gossypium hirsutum L., Malvaceae, America), and Abroma augusta (L.) L.f. (Sterculiaceae, Southeast Asia). A short branch or unbranched mutation was described by Patel, Ghose, & Sanyal (1945) in Corchorus, and by Kearney (1930) in cotton. Both mutations belong to Corner's model (Fig. 8.5E). Another mutation, "abortive terminal" (Fig. 8.5F), was found in cotton and carefully studied by Quisenberry & Kohel (1971); nearly the same mutation exists in Abroma, and was found in the Frère Gillet Botanical Garden, Kisantu, Zaire, in 1969 (Fig. 8.5G). The abortive terminal mutation occurs in many herbaceous plants such as Mollugo, Gisekia, Glinus, Solanum, Euphorbia, and Boerhaavia. It may have played an important part in the process of miniaturization (Hallé & Oldeman, 1975) or evolution from trees to herbs. In Table 8.1 are listed most of the examples of architectural mutations known at the present. The profound heterogeneity of these examples must be stressed. A number of internal or environmental conditions may give rise to these mutations: 1. Profoundly modified ecologic or experimental environment (apical sexuality in Ravenala, plagiotropy in Paliurus) 2. Surgery or chemical mutagen experiment (apical sexuality in Impatiens, dichotomous branching in maize) 3. Viruses or mycoplasmas (forma monstrosa in Opuntia) The genetic significance of these architectural mutations is also variable. In a few examples, the genetic constitution of the species is not modified, and the mutation is either not heritable (lampbrush in Hevea) or its heritability is not clearly demonstrated (foxtail in Pinus). In some cases, the mutation is genetic and involves a single nuclear gene (indeterminate growth in maize, short branch in cotton, lack of branching in jute and sunflower). In these cases, the mutant form of the gene is recessive, and the inheritance is a very simple Mendelian one. On the other hand, the abortive terminal mutation in cotton, studied by Quisenberry & Kohel (1971), is not under the control of nuclear genes and has proved to be cytoplasmically inherited through the maternal parent. This result must be compared with those of Demarly (1974), who obtained, in Lactuca plants of equal genotypic constitution, several forms of branching (distal branching, basal branching, no branching). In Demarly's experiment, the different plants originated from the same tissue culture, and the equality of the genetic constitutions was easily demonstrated by crossing. The results in cotton and lettuce suggest that cytoplasmic heredity plays an important part in the genetic control of the vegetative architecture. Table 8.1. Known architectural mutants. Conclusions. In spite of the considerable diversity of architectural mutations, this survey leads to a preliminary conclusion that seems to be of importance. In most cases, the form of the mutant is not arbitrary; it belongs to a known model, one which is normal in other species. It was recently assumed by Meyen (1973) and Stone (1975) that plant morphology is a closed system and that species must conform, over and over again, to a few basic designs. Some experimental results (Nanda & Purohit, 1966; Roux, 1968; Hallé & Martin, 1968; Mouli, 1970) now support the concept of plant architecture as a closed system. Stevens (1974) suggests that the concept of a closed system may apply to all sorts of morphologic patterns, irrespective of whether they are living or inanimate. Future research on tree architecture and its genetic background will depend upon knowledge of other examples of architectural mutations, and I would appreciate receiving information on the subject based on either field observation or published report. The possibility of crossing an architectural mutation with its normal form provides a method for genetic analysis of vegetative architecture and leads ultimately to the causal phylogeny of tree models. General discussion. Tomlinson: The mutation of Ravenala is interesting because the mutation is in the direction of the genus Phenakospermum, of the same family. Phenakospermum is distinctive in the family Strelitziaceae in that it propagates by basal stoloniferous suckers, and there is a very interesting correlation between the sexuality of the axis and the method of vegetative growth. Would you speculate that Ravenala that did not have basal suckers would be poorly adapted and would not survive? Hallé: I don't know, because all the apical flowering Ravenala I saw were in cultivation and suckers were removed for propagation purposes. Tomlinson: Is there a possibility that this manipulation was the stimulus for the flowering? Hallé: No, I don't think so. Janzen: I would like to know what is the sex determination mechanism in cycads? You showed male and female plants. Does that start out as a 50: 50 sex ratio? Or is sexuality determined later in the life of the plant, which then remains either a male or female? Hallé: If I remember correctly, I found many more males than females. I don't know if the genetic mechanics of sex determination are known in this particular Cycas species. Alvim: I would like to ask Dr. Hallé about the change in the architecture of Theobroma cacao (Nozeran's model). In some areas the plagiotropic branches are almost orthotropic, although they retain the distichous phyllotaxis; they are very thick and vigorous. I wonder how would you interpret this? Is it a mutation? Hallé: Obviously I cannot answer without seeing the plant, but I would be very interested to know the mechanism of growth of the trunk. Is it the same sympodial mechanism of normal cocoa, or does a single meristem continue to grow? Alvim: The first orthotropic shoots produce a crown and plagiotropic branches. Subsequently, another upright, very thick axis is developed from one of the plagiotropic branches. This retains the same phyllotaxis of the plagiotropic branch, but lacks a plagiotropic orientation. They are essentially orthotropic and very vigorous. I shall try to propagate or graft it to see if it retains this character. It is certainly associated with some environmental stimulus because it is found only in a few areas where the conditions must be different; but only 1% or 2% of the plants show the change. Oldeman: The axes you describe are a reiteration of the model. If this is true, they represent rejuvenation in the tree and you might well be able to make very good cuttings of them. They are related to increased energy levels in the environment because reiteration is always opportunistic. Borchert: In relation to Dr. Hallé's and Dr. Alvim's statements, can one always assume that the observed morphologic variation is genotypic and not phenotypic? In your own example of the transition of Hevea from rhythmic to continuous growth and vice versa, you have expressed within the same plant two models, so obviously within one plant of the same genotype both behavioral forms are possible. Likewise, foxtailing is never observed in the temperate zone, yet many temperate zone Pinus when grown in the tropics (e.g., Colombia) for lumber production, grow foxtails. Some of the California pines do it consistently; this is one reason they have not been successfully used for timber production in the tropics. So again we must have with the same genotypic basis different phenotypic expressions. Can one easily refer to all of these as genetic mutations? Hallé: I agree with you completely. This is why I put the term "mutation" in quotation marks. Whitmore: In a young field of rubber planted in Malaya, some of them foxtail and others do not, yet all are the same clone. Givnish: Have you looked at a few families that have a diversity in architectural models and considered how there may be an adaptive radiation within the family accordingly? Hallé: This is a very important question, but the answer is not easy. For instance, in the boundary between forest and savanna, you find mainly an architecture with high potential for vegetative propagation. Otherwise, there is no clear evidence of a relation between architecture and ecology. Oldeman: I do not quite agree. Certainly the relation between architecture and ecology is rather loose. On the other hand, when a tree conforms to the model throughout its whole life, it generally grows in an environment where energy is fairly constant. Such trees are uncommon in the temperate region; nor does one find such trees in the "mature" rain forest canopy, because most of the canopy trees have grown up from the shade into full light - i.e., they have experienced an appreciable change of energy offered during their lifetimes. References. Carvalho, A., & An tunes Filho, H. (1952). Novas observacoes sobre o dimorfismo dos ramos em Coffea arabica L. Bragantia, 12, 81-4. Chamberlain, C. J. (1911). The adult cycad trunk. Bot. Gaz., 52(2), 81-104. Cours, G. (1951). Le manioc à Madagascar. Mém. Inst. Sci. Madagascar B, 3, 203-400. Cremers, G. (1975). Sur la présence de dix modèles d'architecture végétative chez les Euphorbes malgaches. C.R. Acad. Sci. Paris, 281, 1578-8. Demarly, Y. (1974). Quelques données nouvelles en amélioration des plantes. Agron. Trop., 24, 9. Dupuy, P., & Guédès, M. (1969). Homologies entre rameaux courts, rameaux longs et rameaux reproducteurs chez le pin maritime (Pinus pinaster Ait. ssp. maritima H. del Vill.): Notion de métamorphose chez les gymnospermes. C. R. Sci. Soc. Biol., 163(11), 2426-9. Dupuy, P., & Guédès, M. (1975). Une précieuse forme du pin maritime: Pinus pinaster Ait. subsp. atlántica H. del Vill. cv. "piboteau." Bull. Soc. Bot. Fr., 122, 163-72. Guinier, P. (1935). Une curieuse mutation chez le sapin (Abies alba): Les sapins sans branches. C. R. Sci. Soc. Biol., 119, 863-4. Hallé, F. (1971). Architecture and growth of tropical trees exemplified by the Euphorbiaceae. Biotropica, 3, 56-62. Hallé, F. (1974). Architecture of trees in the rain forest of Morobe District, New Guinea. Biotropica, 6, 43-50. Hallé, F., & Martin, R. (1968). Etude de la croissance rythmique chez l'Hévéa (Hevea brasilienáis Müll.-Arg. Euphorbiacee-Crotonoïdée). Adansonia, 8, 475-503. Hallé, F., & Oldeman, R. A. A. (1975). Essay on the Architecture and Dynamics of Growth of Tropical Trees. Tr. B. C. Stone. Kuala Lumpur: Penerbit Universiti Malaya. Hallé, N. (1971). Crioceras dipladeniiflorus (Stapf.) K. Schumann. Apôcynacée du Gabon et du Congo. Adansonia, 11, 301-8. Hallé, N. (1973). Captaincookia, genre nouveau monotypique néo-calédonien de Rubiacéae-Ixoreae. Adansonia, 13, 195-202. Henry, P. (1948). Un Elaeis remarquable: Le palmier à huile vivipare. Rev. Bot. Appl, 311, 312, 422-7. Henry, P., & Scheidecker, D. (1953). Nouvelle contribution à l'étude des Elaeis vivipares. Oléagineux, 10, 681-8. Hockett, E. A., & Knowles, P. F. (1970). Inheritance of branching in sun.flowers, Helianthus annuus L. Crop Sci., 10, 432-6. Jones, D. F. (1921). The indeterminate growth factor in tobacco and its ef.fects upon development. Genetics, 6, 433-44. Kahn, S. (1975). Remarques sur l'architecture végétative dans ses rapports avec la systématique et la biogéographie. D.E.A. Biol. Vég., Montpellier. 33 pp. Kearney, T. H. (1930). Short branch: Another character of cotton showing monohybrid inheritance. /. Agrie. Res., 41, 379-87. Kozlowski, T. T., & Greathouse, T. E. (1970). Shoot growth and form of pines in the tropics. Unasylva, 24, 2-10. Kummerov. J. (1962). Ueber Wachstumsanomalien bei Pinus radiata unter tropischen Bedingungen. Ber. Dtsch. Bot. Ges., 75, 37—40. Lewalle, J. (1968). A note on Hyphaene ventricosa. Principes, 12, 104-5. Mabberley, D. J. (1973). Evolution in the giant groundsels. Kew Bull., 28, 61-96. Mabberley, D. J. (1974). The pachycaul Lobelia of Africa and St. Helena. Kew Bull., 29, 535-84. Mabberley, D. J. (1975). The pachycaul Senecio species of St. Helena, "Cacalia paterna" and "Cacalia materna." Kew Bull., 30, 413-20. Meyen, S. V. (1973). Plant morphology in its nomothetical aspects. Bot. Rev., 39, 205-60. Mouli, C. (1970). Mutagen-induced dichotomous branching in maize. /. Hered., 61, 150. Nanda, K. K., & Purohit, A. N. (1966). Experimental induction of apical flowering in indeterminate plant Impatiens balsamina L. Naturwissenschaften, 54 , 230. Nozeran, R., & Neville, P. (1963). Nature "virale" de certaines transformations observées chez des Cactacées. Nat. Monsp., 15, 109-14. Patel, J. S., Ghose, R. L. M., & Sanyal, A. (1945). The genetics of Corchorus (jute). III. The inheritance of corolla colour, branching habit, stipule character and seed coat colour. Ind. Genet. Plant Breeding, 4, 75-9. Quisenberry, J. E., & Kohel, R. J. (1971). Abortive terminal, a cytoplasmically inherited character in cotton, Gossypium hirsutum L. Crop Sci., 11, 128-9. Reeves, R. G., & Stansel, R. H. (1940). Uncontrolled vegetative development in maize and teosinte. Amer. Bot., 27, 27-30. Roux, J. (1968). Sur le comportement des axes aériens chez quelques plantes à rameaux végétatifs polymorphes: Le concept de rameau plagiotrope. Ann. Sci. Nat. (Bot.), 12, 109-255. Scarrone, F. (1969). Recherches sur les rythmes de croissance du manguier et de quelques végétaux ligneux malagasy. Thesis, Université de Clermont-Ferrand. 2 vols. Singleton, W. R. (1946). Inheritance of indeterminate growth in maize. Hered., 37, 61-4. Soria, V. J. (1961). A note on cacao plant which flowered at the age of three months. Cacao, 6, 11-12. Stevens, P. S. (1974). Patterns in Nature. Boston: Atlantic Monthly Press. Stone, B. C. (1975). Translator's preface. In Essay on the Architecture and Dynamics of Growth of Tropical Trees, by F. Hallé, & R. A. A. Oldeman. Kuala Lumpur: Penerbit Universiti Malaya.