Acta Botanica Neerlandica 15 (1966) 1-33 Remarks on the position, the delimitation and the subdivision of the Rubiaceae C.E.B. Bremekamp (Botanical Museum and Herbarium, Utrecht) (received October 22nd, 1965) Introduction It is often assumed that the delimitation and the subdivision of the various families which have been distinguished in the Angiosperms, do offer serious no longer difficulties. They would belong to those of for which objects study already long ago a fairly satisfactory so- found. If wish be with this lution was we to acquainted solution, the would have would be look such works only thing we to do, to up as Benthamand Hooker’s “Genera Plantarum” and Engler and Brand's “Nattirliche Pflanzenfamilien”. Some improvements might still be but these would be of minor desirable, importance only. These as- be sumptions, however, are to regarded as dangerous illusions. That the very serious nature of the shortcomings found in the delimitation and subdivision of these families, especially of the larger ones, is so often overlooked, is apparently due to an attitude of mind which is observed in of a comparatively large part the taxonomists, viz. a lack of interest in the development of a truly natural classi- fication. This is not incomprehensible. Most of them spend the major part of their time in the elaboration of floras covering areas of more but of or less limited extent, and they are rarely aware the fact that the the which is in knowledge of families obtained this way, remains in necessarily incomplete. Moreover, the elaboration of a flora the most essential point is the construction of serviceable keys to the species as well as to the groups of higher rank, not the exact deli- mitation of these the latter end material is groups; to usually more than the of flora has required compiler a at his disposition. However, we must not overlook the fact that a key, in order to be serviceable, need not reflect the degree of affinity between the units with which it is dealing; in reality, such keys are often entirely or almost entirely this therefore artificial, and applies also to classifications which are will of based on such keys. To illustrate this, I discuss here some the aid of which in the Rubiaceae suitable the characters by very keys to the genera have been constructed, but which when they subse- used the quently were for elaboration of a classification, led to entirely unsatisfactory results. time have To the characters which for a long played an important either part in the subdivision of the Rubiaceae, belong the presence of their kind of which to of uni- or pluriovular ovary cells, fruits, 1 2 C. E. B. BREMEKAMP this end divided in and and the of their were dry fleshy ones, nature which viz. seeds, of also two kinds were distinguished, winged and of the that wingless ones. However, even a superficial study groups have been distinguished by the aid of these characters, will show that As most of them are entirely unnatural. the differences are morpho- of this have been logically hardly any value, might expected. The number of ovules cell shows continuous of per ovary a range there variation, and so long as is no reason to assume a correlation between a definite number of ovules and one or more other characters, it is therefore attach the of not justified to to presence one or two ovules cell taxonomic value per ovary a greater than to that of three or of another number of them; we know, in fact, that the number of cell within ovules per ovary may vary the limits of a single genus of the Pavetta L” in (see my “Monograph genus Fedde, Repertorium 37: remarks 10. 1943, and my on the genus Anotis DC and the in the of the Cruckshanksieae last chapter present publication) as well as in genera that on account of their many common features are to be in Tarenna Gaertn. and Pavetta regarded as nearly related, so e.g. L (see p. 7 of my “Monograph of the genus Pavetta L”). the between and be That difference dry fruits fleshy ones can not the of used for characterization natural groups is also easily com- This distinction is prehensible. an ecological, not a morphological one, and in both kinds of fruits quite considerable morphological differ- ences are found; in some instances the differences between a dry and than those between kinds of a fleshy fruit may even be smaller two fruits kinds of dry or two fleshy ones. In the genus Mussaenda L species with dry fruits are found side by side with species producing if fleshy ones; even these species are divided over two genera, as sometimes is done, this does not make much difference, for in that case we will have to admit that the two kinds of fruits are found in related very nearly genera. The same applies to the distinction between winged and wingless seeds: this is too an ecological, not a morphological distinction. From the morphological point of view these wings are by no means all alike; in the the tribe in which the with seeds Cinchoneae, genera winged were in the which be brought together, they are genera may rightly regarded as near allies of Cinchona L, in position as well as in the structure of the cells of which they are mainly composed, entirely different from those found in which that are the genera were trans- ferred by me to the Rubioideae. Another drawback of the classifications accepted in the works is mentioned above, that they are based mainly on earlier classifi- which cations were drawn up when a much smaller number of genera were known, and that the authors often assumed that the subfamilies and even the tribes based on this more limited number would suffice for the of the discovered This incorporation subsequently genera. ap- is the of the that parently explanation fact so often a genus has been In the the squeezed into a tribe into which it does not fit. case of Rubiaceae there are even quite a number of genera of which it is POSITION, DELIMITATION AND SUBDIVISION OF THE RUBIACEAE 3 extremely doubtful whether they really belong to this family. This point will be discussed in the second chapter. Many taxonomists, moreover, seem to feel a certain reluctance to subfamilies which contain accept but a single tribe, and tribes which contain but a single genus. This reluctance, however, is entirely We unjustified. know several families and even some orders which contain but a or a small number of and there is single very genera, that accordingly no reason whatever for assuming a subfamily or a tribe should contain than of the always more one group next lower rank. Whether is a taxon to be classified as a subfamily or a tribe evidently depends upon the numberand the importance of the charac- in it differs ters which from its nearest allies and on nothing else, and where our haveoverlooked the of predecessors presence some differences undervalued their is or importance, it doubtless our duty tocorrect them. With regard to what has been said in the preceding paragraph, it of of for is, course, importance to find ways and means estimating the taxonomic value of the of groups characters that are to be used for the distinction of the tribes and subfamilies and, naturally, of the In the of families too. case the latter the number of diagnostic charac- is often ters extremely small and their value not rarely dubious. With regard to the Rubiaceae this point will be dealt with in the chapters and one two, but at this point it is perhaps worth while to draw the attention to the fact that in the description of this family given in the works mentioned above we do not find a single character that can be as This of regarded a truly general one. is, course, due to the fact that this family, in the delimitation accepted in these works, elements comprises which do not really belong to it. The inclusion of such elements, moreover, has often led to entirely false conclusions with regard to the affinity of the families. In the case of the Rubiaceae, for the inclusion of instance, the genera Henriquezia Spruce ex Bth. and Platycarpum Humb. et Bonpl. has led to the assumption that there should be a comparatively close affinity between this family and the However, when these two are it Bignoniaceae. genera excluded, ap- pears that not a single argument is left which might possibly be adduced in favour of this view. After the exclusion of the genera Carlemannia Bth. and Sylvianthus Hook. f. the affinity with the Capri- and their allies has become less and if the foliaceae pronounced, genus Dialypetalanthus Kuhlm. had been left in the Rubiaceae, this would have been rather a good argument for assuming a close affinity with the ! It Myrtales is therefore of great importance that such misclassifications corrected. Several other of are examples genera which were incorpo- ~ ~ rated in the Rubiaceae on insufficients ground, will be given in the chapter dealing with the delimitation of the family. It of be denied can, course, not that the number of characters which related taxa have in decreases common, with the increase in rank of the taxon to which they belong; it is certainly no wonder that families but have, as a rule, a very small number of general characters. However, groups based on a single general character should always be regarded with some distrust. 4 C.