New PhytoL (1976), 76, 551-554. A NOTE ON THE EVOLUTIONARY STATUS OF ALUMINIUM-ACCUMULATORS AMONG DICOTYLEDONS BY ERNEST M. CHENERY* AND KENNETH R. SPORNE Colinas Verdes, Lagos, Portugal, and Tiie Botany School, University of Cambridge {Received'] October 1975) SUMMARY Among dicotyledons, the ability to accumulate large quantities of aluminium is statistically correlated with seven primitive characters. It is especially common among those families that were present in the early fossil history of the group and among tropical rain-forest families. It is concluded, therefore, that aluminium-accumulation is itself a primitive character. The leaves of most plants contain only minute amounts of aluminium (e.g. less than 300 ppm/dry weight) but some may contain larger amounts (more than 1000 ppm). Such plants are known as 'aluminium plants', or 'aluminium-accumulators'. They have been recorded by one of us (Chenery, 1948, i95ia,b) and by Webb (1954) in thirty-four of the families of dicotyledons recognized by Diels (1936). Recent work (by E. M. C.) has completed the investigation of representatives of all the families and has increased to thirty-seven the number in which aluminium-accumulators have been recorded. They are listed in Table i, where they are arranged in the Orders recognized by Diels. It is elear from Table i that aluminium-accumulators are widely scattered among the major subdivisions of the dicotyledons. Above the family level, therefore, aluminiumaccumulation has little taxonomic significance. However, it has been pointed out (Chenery, 1948) that, with very few exceptions, aluminium-accumulators are arborescent. This was, naturally, the first feature that led us, like Webb (1954), to think that aluminium-accumulation might be associated with primitiveness. Statistical analysis (by K. R. S.) confirms that this is so. In Table i, alongside each of the families there is a number corresponding to its percentage advancement index (taken from Sporne, 1969). The average for the thirtyseven families is 49.0, whereas that for the families examined for aluminium-accumulation is 56.6. These two averages are significantly different, as is shown by applying a statistical test appropriate to skew distribution curves (Sporne, 1970), for the value of P = 0.0054 obtained implies a i :2i6 chance that the difference is fortuitous. Such a difference between the average advancement indices leads one to expect that aluminium-accumulation may be statistically correlated with at least some of the twentyfour primitive characters listed by Sporne (1969), for twenty-two of those characters were used in calculating the advancement indices. The application of two-by-two contingency tests to the data shows that aluminium-accumulation is indeed positively * Formerly at the Tea Research Institute, Ceylon.
552 E. M. CHENERY AND K. R. SPORNE Table i. Orders and families of dicotyledons, some members of which are aluminium-accumulators. The advancement index of each family is shown as a percentage. Those marked with an asterisk are recent additions, hitherto unpublished. Juglandales Juglandaceae Fagales Fagaceae* 29 Proteales Proteaceae 50 Santalales Octoknemataceae Olacaceae 63 Ranales Lardizabalaceae* Lauraceae 42 Magnoliaceae 27 Monimiaceae 43 Rosales 62 Cunoniaceae 24 Saxifragaceae 45 Geraniales Dapbnipbyllaceae 53 Eupborbiaceae 30 Polygalaceae 63 Rutaceae 50 Vocbysiaceae 62 Sapindales Celastraceae 48 Icacinaceae 47 Pentapbylacaceae 31 Malvales Scytopetalaceae 53 Parietales Flacourtiaceae* 22 Theaceae 32 Violaceae 42 Myrtiflorae Combretaceae 45 Crypteroniaceae 72 Geissolomataceae 64 Melastomataceae 45 Myrtaceae 27 Rhizophoraceae 21 Umbelliflorae Cornaceae 55 Diapensiales Diapensiaceae 75 Ebenales Diclidantheraceae 73 Symplocaceae 53 Contortae Apocynaceae 57 Gentianaceae 76 Loganiaceae 75 Rubiales Rubiaceae 70 Table 2. Correlations between aluminium-accumulation and various primitive characters, occurrence in the fossil record and occurrence in tropical environments Cbaracter 'Y' 1, Woody babit 2, Vessels with scalai'iform end-plates 3, Vessels witb scalariform side-walls 4, Wood witb apotracbeal parencbyma (cf, woody families) 5, Unstoreyed wood (cf, woody families) 6, Leuco-antbocyanins present 7, Integument vascular bundles present 8, Pre-Oligocene record (cf, modern world) 9, Pre-Tertiary record (cf, modern world) 10, Pre-Tertiary record (cf, complete fossil record) 11, Pre-Turonian record (cf, modern world) 12, Tropics (cf, whole world) 13, Rain-forest (cf, whole world) 14, Rain-forest (cf. Tropics) X 29 30 26 n = tbe number of families about wbicb the facts are known, in eacb case; x = tbe number of families witb aluminium-accumulating members; y = tbe number of families witb cbaracter 'Y'; m = tbe number of aluminium-accumulating families wbicb also exbibit cbaracter 'Y'\ m' = -^ = tbe number of alun minium-accumulating families that would be expected, on a null bypotbesis, to exbibit character 'Y'. The symbol + indicates a positive correlation at a 50:1 level of significance, i,e, wbere ;;;^>5,4, The symbol + indicates one at a 20:1 level, i,e, where 3,8<;i;^< 5.4), t Yates' correction for small numbers was applied. y 216 99 43 JIO 141 98 68 58 58 45 192 71 71 n 230 203 178 195 187 224 138 192 m 28 15 27 31 23 15 10 16 16 12 20 20 ni 30,8 15,9 7-4 20,4 25,3 15,2 9,1 5,0 8,3 10,9 6,4 27,4 10,1 12,2 X Correlation 8,4 + 19,2 11,9 + -1-6,9 + 5-6 + 9,9 + 6,3 + 5,5t + 10,7 -h 5.1 + 6,8 + 5,1 + 15-6 + 9,6 +
Aluminium-accumulators 553 correlated with seven primitive characters at the 50: i level of significance. The facts on which these calculations are based are displayed in Table 2. There can be little doubt that the twenty-four characters listed by Sporne (1969) are primitive, not only because they are statistically associated with one another, but also because they are relatively more abundant in those families which appeared earliest in the history of dicotyledon evolution. Not only is aluminium accumulation associated with woodiness (correlation number i, in Table 2) but it is also correlated with primitive types of wood, within woody families (correlations 4 and 5). Furthermore, it is correlated with those kinds of vessel-element that have been universally accepted as primitive for more than 50 years (correlations 2 and 3). It seems clear, therefore, that aluminiumaccumulation is, itself, a primitive character. This is confirmed by correlations 8-11. The first of these shows that those families recorded by Muller (1970) as having been in existence by the start ofthe Oligocene epoch have a higher proportion of aluminium-accumulators than the present-day fiora of the world. Correlation 9 shows that the same is true of those families recorded by Chesters Gnauck and Hughes (1967) as having been in existence by the start of the Tertiary period. Correlation 10 shows that these same pre-tertiary families have a higher proportion of aluminium-accumulators than do families recorded from younger fossil deposits. Correlation 11 shows the result of comparing those families listed by Chesters et al., as having been in existence by the end of the Cenomanian epoch with the presentday flora of the world. Again, there is a significant difference in the proportion of aluminium-accumulators. Almost throughout the entire history of the dicotyledons, therefore, there have been higher proportions in the past than there are at the present day. It has been shown (Sporne, 1973) that primitive characters occur more frequently in tropical rain-forests than anywhere else in the world, and this is true also of aluminiumaccumulation, as is shown by correlations 12, 13 and 14. In the first of these, the consolidated floras of French Guiana, West Tropical Africa, Tropical Africa, the Malay Peninsula, Java, Samoa, South-east Polynesia, New Caledonia and the Philippines (Sporne, 1973) are compared with the flora of the world. In correlations 13 and 14, consolidated lists of families from three rain-forest localities in Guiana, Uganda and Sarawak are compared with the flora of the world and with that ofthe tropics, respectively. It is well known that highly leached tropical soils contain large quantities of available or exchangeable and ionic aluminium; and the importance of this fact to tropical ecologists has been emphasized (Chenery, 1948). It is now seen to be of importance, also, to students of flowering plant evolution. In order to survive in most tropical rain-forests, plants must either be able to tolerate the presence of aluminium in their tissues or must have developed some mechanism for excluding it. Primitive dicotyledons are characterized by the first of these alternatives, for any plant which can build up a content of aluminium of 1000 ppm in its leaves must surely be described as tolerant. ACKNOWLEDGMENTS One of us (E. M. C.) is grateful for the stimulating encouragement that he received from Dr A. Cronquist, Mr R. C. Barneby and Dr R. J. A. Goodland of the New York Botanical Garden and to Dr P. C. DeKock of the Macaulay Institute for Soil Research, Aberdeen. Sincere thanks are due to Mr J. P. M. Brenan, Deputy Director, Royal
554 E. M. CHENERY AND K. R. SPORNE Botanic Gardens, Kew and Dr P. K. Holmgren of New York Botanical Garden for the privilege of testing herbarium specimens under their charge. REFERENCES CHENERY, E. M. (1948). Aluminium in the plant world. Part i, General survey in dieotyledons. Kew BuU., CHENERY,' E. M. (1951a). Contributions to the biogeochemistry of aluminium, 1948-1950. Colonial Office Paper, 1529. CHENERY, E. M. (195ib). Some aspects of the aluminium cycle. J. Soil Sci., 2, 97. CHESTERS, K. I. M., GNAUCK, F. R. & HUGHES, N. F. (1967). Angiosperms. In: The Fossil Record (Ed. by W. B. Harland et al.), p. 269. Geology Society, London. DIELS, L. (1936). A. Engler's Syllabus der Pfianzenfamilien. Borntraeger, Berlin. MULLER, J (1970). Palynological evidence on early differentiation of of angiosperms. Biol. Rev., 45, 417. SPORNE, K. R. (1969). The ovule as an indicator of evolutionary status in angiosperms. New PhytoL, 68, SPORNE, K. R. (^1970). The advancement index and tropical rain-forest. New PhytoL, 69, 1161. SPORNE^ K. R. (1973). The survival of archaic dicotyledons in tropical rain-forests. New PhytoL, 72, 1175. WEBB, L. J. (1954). Aluminium accumulation in the Australian-New Guinea flora. Aust.J. Bot., 2, 176. 6