Origin of angiosperm characters

Valentin A. Krassilov

Krassilov VA 1996. Origin of angiosperm characters. Palaeobotanist45: 400-406.

Models of origin for the typical angiosperm leaf, flower, stamen, pistil and double fertilization are based on evolutionary trends in proangiosperms. It is suggested that angiosperm organs are of chimeric origin, acquired by aggregation and fusion of progenitorial structures. J'hese morphological processes might involve different proangiosperm lineages in unstable (ecotonal, tectonically active) environments. An advantage of early angiosperms in such environments might be due to extended evolutionary potentials of their chimeric organs capable of acquiring new functions related to entomophily and zoochory.

Key-words----Angiosperm origin, Morphology, Evolution, Proangiosperms.

Valentin A. Krassilov, Palaeontological Institute, 123 Proftojusnaya, Moscow 117647, Russia.

~a:i1~aan-il~. ~. ~. '1fTlffrrv:ctft~tri~am:r-a:i1~aan-iltilmcfilFC1Cfi1mii ~ tR" ~ ~I ~ stfdlfqd ~ TfllT l ~ a:tl~do:ft\Jll WT ~ ~· tf; l"lfat"1Uf OITR CfiT ~ ~I ~ a:tiCfilRil~daft;ffi4iit~Cfif4~~~-~

ANGIOSPERMS are plants of considerable structural ideal, or archetypal, rather than typical, for actually complexity including a number of characters, i.e., there could be none of the kind growing around. structural features essential for their description as a Even the expectations of finding archetypal an­ category of plant kingdom. There are characters they giosperm as a fossil are rapidly expiring with the share as well as those they do not share (or if share progress of palaeobotanical research. then reluctantly) with other plant categories. Among It is, thus, clear that the notion of angiosperm the non-shared characters, some are present in all or characters is meaningful only in comparison with majority of the hitherto studied angiosperms, the another plant category showing a number of similar It is character states alongside some dissimilar char­ the virtue which this latter category Our current no·- KRASSILOV-ORIGIN OF ANGIOSPERM CHARACTERS 401

PIATE 1 402 TIIE PALAEOBOTANIST

angiosperms were taxonomically separated from giosperms, thus providing an angiosperm character gymnosperms. The mid-XIX century system-makers pool (Krassilov, 1975). Whether they all or any of have treated the both as divisions of Anthophyta, them actually gave rise to plants currently classified flowering plants, thus using the naked vs protected as angiosperms is another matter. More angiosperm ovules, and not the flower, as a principal distinction. characters are related to evolutionary trends in proan­ As far as the extant plants are concerned (and giosperms the more likely the angiosperm ancestry there was nothing else to be concerned at the time) of the latter would seem. Here I briefly summarize the the distinction is clear cut. With addition of fossil character origin models described in more details plants, such as cupuliferous pteridosperms or elsewhere (Krassilov, 1977, 1989, 1991): caytonias, it is not. On the other hand, the much 1. Proangiosperms including dirhopalostachians, debated problem whether the conifer seed-scale leptostrobaleans (czekanowskialeans), cayto­ complexes are flowers (of the same kind as the seed­ nias, and extinct gnetaleans were represented by scale complexes of grasses) or not seemed to have leptocaul trees with deciduous leafy spur shoots, been resolved by Florins (1938-1945) work on fossil shrubs and waterside - aquatic or semiaquatic - conifers showing that their seed-scales are herbs (Baisia: Krassilov & Bugdaeva, 1981). All ovuliferous shoots rather than ovuliferous leaves as these life forms might occur in early angiosperms. angiosperm carpels were then supposed to be. In Incidentally, Caspiocarpus, an early Albian an­ effect flowers were rendered typical of angiosperms giosperm from Kazakhstan with pistillate making synonymous to flowering plants. panicles attached to leafy shoots (Vachrameev & Flower itself was, with admirable circularity, defined Krassilov, 1979; Krassilov, 1991) was recons­ as a reproductive organ of flowering plants, although tructed as a herbaceous plant, and a quite similar bennettitalean flowers still posed some problems. form was then described from Koonwarra bed in when fossil data infringe a taxonomic Australia (Drinnan & Chambers, 1986). distinction the latter can nevertheless be sustained 2. The monocot-type parallel-veined leaves might either by ignoring the fossil data or by claiming the have phyllode origin (Arber, 1918) as evidenced fossil structures non-homologous - and therefore ir­ by histological studies (Kaplan, 1970) and similar relevant - to morphologically similar extant struc­ tendency in czekanowskians, gnetaleans ( tures, and both the methods have been vigorously witschia) and bennettites. Some of the by plant morphologists. Even aberrant phyllodial bracts still bearing miniature leaf living plants, such as gnetaleans, were treated in the blades (Harris, 1969; Krassilov, 1982a). same way. The morphological proximity of Cretaceous angiosperm leaves ly branching segments (Debeya, Proteophyllum, etc.), as well as Scoresbya-Sagenopteris leaf show inter- mediate shoot/leaf (caulome/ phyllome) ters (Krassilov, 1989). The typical dicot-type blade origin is modelled on the basis of segment fusion in compound leaves of peltasperms in­ cluding such supposedly proangiospermous plants as Furcula. Segment fusion by marginal meristems initially has not affected the originally open venation pattern. However, the

PI.ATE 2 l. Pistillate head from the Palaeocene: of Kamchatka, x 20 (Maslova 4. Lemnoid fruit from the Maastrichtian of Mongolia. SEM, x 70 & Krassilov. in press). (Krassilov & Makulbekov, in press). 2. Individual flower of the same, SEM, x 30. 5. Ovule of the same showing funicle and embryo cap, SEM, x 10. 3. Flower showing distinct staminodes (s) and interfloral phyllomes (i), SEM, x 30. 6. Funicle of the same, SEM, 660. KRASSILOV-ORIGIN OF ANGIOSPERM CHARACTERS 403

PIATE 2 404 THE PALAEOBOTANIST

hopalostachians. Some early angiospem flowers still retained a little modified anthocorm mor­ phology (Dilcher, 1979). Moreover, distinct inter­ floral phyllomes were recently found in florets of Palaeocene hamamelid heads (Maslova & Kras­

A silov, in press; Pl. 2, figs 1-3) indicating their primary rather than derived character in an­ giosperms. 5. Stamen origin is modelled on Meeusella, a staminate shoot with lateral branches bearing a pair of stalked sporangial heads (Krassilov & Bugdaeva, 1988; Pl. 1, fig. 3). Some lateral branches are sterile. The sporangial stalks­ second order branches, are of variable length along the axis. In the case of their extreme reduc­ tion the paired sporangial heads are sessile, with the primary branch apex protruding between them as a connective of bithecate anther. The latter thus might derive from a pair of second E order sporangiate branches while reduction of Text-figure I-Reconstruction of proangiospenn cupules: A- Lep­ the main axis would give a fascicle of stamens. tostrobus, B-Dtnopbyton, C-Eoantha, D-Batsia,. E-Caytonta and This model implies primary nature of fasciculate putative prototype with distinct subtending bract (after Krassilov, stamens (cf. Pauze & Sattler, 1978) as well as 1969). paired stamens, as in Saururaceae (Tucker, 1985) meristem occurring within the aggregate leaf and possibly also stamens with sterile ap­ blade might give rise to plate meristem which pendages (derived from sterile lateral branches mediated looping of the former midveins and, of Meeusella prototype), as in Monimiaceae and within the primary loops, oflateral veins and their Lauraceae. branches, thus forming areolate venation of 6. The tricolpate and triporate pollen grains appear­ several orders (Krassilov, 1991, 1995; Plate 1). ing rather early in the fossil record have no ob­ 3. The long debated homology of vessels in vious precursors among proangiosperms. How­ gnetaleans and angiosperms has been convinc­ ever, the monosulcate fossil pollen grains of ingly demons-trated by Muhammed and Sattler Eucommiiditestype have a zonal equatorial fur­ (1982). In addition, tracheid-like vessels with row (Hughes, 1994) or two additional furrows simple porous and scalariform perforation plates parallel to the sulcus. These furrows are scars of were found in bennettitalean leaf veins (Kras­ reduced sacci of monosaccate or bisaccate pollen silov, 1982a) evidencing the occurrence of this grains occurring in .caytonialeans and other character in more than one proangiosperm proangiosperms. A polarity change along is re­ groups. quired to transform Eucommiidites into tricolpate 4. Anthocorm model of angiosperm flower pollen grain of primitive angiosperms. Porate or (Meeuse, 1975) is supported by the occurrence of pseudoporate appertures first appeared in Clas­ anthocorms--essentially short shoots with apical sopollis, fossil pollen traditionally assigned to dusters of pistillate or staminate organs or both conifers but referred to extinct gnetaleans by mixed with interfloral phyllomes ( sterile leaves Krassilov Cl982b). Recently clumps of Classopol­ or cataphylls">--:-in several proangiosperm lis have been found in the guts ofJurassic insects groups, such as Irania, bennettites, czekano­ (Pl. 1, figs 4-6) evidencing pollinivory and possib­ wskias and possibly also caytonialeans and dir- ly entomophily. In these pollen grains the pore KRASSILOV-ORIGIN OF ANGIOSPERM CHARACTERS 405

area is surrounded by subequatorial rimula. They loosely defined proangiospermous groups, Lep­ resemble tetrads of early angiosperms covered tostrobus-Caytonia and Eoantha-Baisia the with common sexine layer and showing annubte cupules of which are comparable with pistils of pores (Walker et al., 1983). Such tetrads might ranalean dicots and monocots respectively. The evolve in triporate pollen grains (one unit lost recenrly found Late Cretaceous utriculate fruits with the change of symmetry) which are then with funnel-shaped stigmas (Pl. 1, fig. 4) contain initially tetrasporic (Krassilov, 1989, 1991). a solitary orthotropous ovule the funicle of which 7. Reinterpretation of typical carpels as peltate or appears as a direct continuation of the pedicel ascidiform strnctures (Rohweder, 1967; Endress, (Krassilov & Makulbekov, in press). The ovules 1983; Erbar, 1986) make it easier to reconcile their show embryo cap and caruncle. These fruits are morphology with their origin from proan­ assigned to aquatic angiosperms related to Lem­ giosperm cupules (Long, 1966; Krassilov, 1977; naceae, a fairly advanced monocot family. How­ Heel, 1981). Among the Mesozoic cupules (Text­ ever, they seem to retain a primitive cupule-like figure 1), the many-seeded bivalved Leptostrobus fruit structure and the ovule type common to a with submarginal anatropous ovules was formed number of monocot families as well as for their by fusion of two open peltasperm-like cupules, protognetalean precursors. the fusion meristem giving rise to stigmatic papil­ 8. Double fertilization in Ephedra involves the egg late crests (Krassilov, 1969, etc.). A similar origin and ventral canal cell (Herzfeld, 1922; Khan, was suggested for the paired Dirhopalostachys 1940). According to the gametophyte cell homol­ cupules (Krassilov, 1975). Caytonia had as­ ogy model proposed by Krassilov (1989) based cidiform cupules a slit-like exostome (mouth) of on Cocucci (1973) the angiosperm embryo sacis which was bordered by a lip-like appendage - a formed of two or several archegonia, the eggs of residual subtending bract. According to this inter­ which function as polar nuclei while the former pretation (Krassilov, 1989) the ovules are ventral canal cell maintains its position close to anatropous while the mouth occurs in the fusion the former neck cells (synergids) in the way of zone of cupule and its subtending bract. The the entering sperm and is sexualized as a secon­ cupule pairing and fusion, or syncupuly, oc­ dary egg (Text-figure 2). curred in different proangiosperms on their way The above models postulate chimeric origin of to aquiring angiospermous carpels. typical angiosperm characters acquired by aggrega­ Another group of proangiosperms had or­ tion and fusion of progenitorial structures. The latter thotropous ovules borne in the four-lobed morphological processes might reflect developmen­ cupules with adnate bracts and tipped with tufts tal acceleration and the ensuing condensation of of hairs, as in Dinophyton (Krassilov &Ash, 1988) developmental events. In my previous publications or with protruding axis crowned with bracts rep­ (Krassilov, 1977, etc.) the process of angiospermiza­ resenting a reduced distal floral node, as in Eoan­ tion was related to enironmental crises. The earliest tha (Krassilov, 1986) or utriculate with macrofossil angiosperm records occur in the Early hairycorona of supposedly similar origin, as in Cretaceous ecotone between the summer-wet Baisia (Krassilov & Bugdaeva, 1982). Among deciduous and summer-dry evergreen zones. them, Dinophyton and Eoantha show definite Ecotonal environments are most sensitive to climatic gnetalean features (including ribbed pollen and related vegetational changes. In addition, such grains in the latter) on the virtue of which they important early angiosperm localities as, Baisa in the are protognetalean as weH as proangiospermous. Lake Baikal region, Koonwarra in Australia and These and perhaps others still undiscovered or Makhtesh Ramon in Israel (Krassilov & Dobruskina, misinterpreted cupules might give rise to dif­ in press) are situated in the then active rift zones. ferent types of carpels while their bracteate or Environmental instability generally promotes ac­ axial appendages might evolve in various stig­ celerated development and phenotypic plasticity. matic structures. Presently we have two rather Therefore, parallel acceleration trends might involve 406 1HE PALAEOBOTANIST different proangiosperm lineages. An advantage of Krassilov VA 1975. Dirhopalostachyaceae - a new family of proan­ giosperms and its bearing on the problem of angiosperm ancestry. early angiosperms in such environments might be Palaeontographica 153B: 100-110. due to extended evolutionary potentials of their Krassilov VA 1977. The origin of angiosperms. Bot. Rev. 43(1): 143-176. chimeric organs capable of acquiring new functions. Krassilov VA 1982a. Early Cretaceous flora of Mongolia. Palaeon­ New findings of proangiospermous pollen in the guts tographica 181B: 1-43. Krassilov VA 1982b. On the ovuliferous organ of Hirmerella. Phyta. D. of Mesozoic insects support the idea of plant-insect D. Pant Comm. Vol.: 141-144. interaction as a factor in the evolution of pollen Krassilov VA 1986. New floral structure from the Lower Cretaceous of seledion on the basis of the password - response Lake Baikal area. Rev. Palaeobot. Palynol. 47: 9-16. Krassilov VA 1989. Origin and early evolution of angiosperms. Nauka, compatibility mechanism involving both stigmatic Moscow (in Russian). and exinal structures. Another factor leading proan­ Krassilov VA 1991. The origin of angiosperms: New and old problems. giosperm cupules to typical angiospermy was endo­ Trends in ecology and evolution 6 : 215-220. Krassilov VA 1995. Scytopbyllum and the origin of angiosperm leaf and exozpochory the role of which considerably characters. Paleont.]. 29(A): 63-75. increased with the rise of mammals. With some luck Krassilov VA & Ash SR 1988. On Dinophyton-protognetalean Mesozoic this scheme will be further tested by the fossil record. plant. Palaeontographica 208B : 33-38. Krassilov VA & Bugdaeva EV 1982. Achene-like fossils from the Lower Cretaceous of the Lake Baikal area. Rev. Palaeobot. Palynol. 36 : REFERENCES 279-295. Krassilov VA & Bugdaeva EV 1988. Meeusella and the origin of stamens. Lethaia 21 : 425-431. Arber A 1918. The phyllode theory of monocotyledonous leaf, with Krassilov VA & Dobruskina IA. Angiosperm fruit from the Lower special reference to anatomical evidence. Ann. Bot. (Gr. Brit.) 32: Cretaceous of Israel and origins in rift valleys. Paleont.]. (in press). 465-501. Krassilov VA & Makulbekov NM . Aquatic plants from the Maastrichtian Cocucci AE 1973. Some suggestions onthe evolution of gametophytes of of Mongolia. Paleont.]. (in press). higher plants. Phytomorphology23(1-2): 109-124. Long AG 1966. Some Lower Carboniferous fructifications from Berwick­ Dilcher DL 1979. Early angiosperm reproduction: an introductory report. shire, together with a theoretical account of the evolution of ovules, Rev. Palaeobot. Palynol. 27: 291-328. cupules and carpels. Trans. R. Soc. Edinburgh 66: 345-375. Drinnan AN & Chambers TC 1986. Flora of the Lower Cretaceous Koon­ Maslova NP & Krassilov VA. New hamamelid infructescences from warra fossil bed (Korumburra Group), South Gippsland, Victoria. Palaeocene of Western Kamchatka. Rev. Palaeobot. Palyno/. (in Mem. Ass. Austr. Palaeontoi. 3: 1-77. press). Endress PK 1983. TI1e early floral development of Austrobaileya. Bot.]b. MuhammedAF & SatterR 1982. Vessel structure of Gnetumand the origin Syst. Pflanzengesch .. Pflanzengeogr. 103(4) : 481-497. of angiosperms. Amer.]. Bot. 69(6) : 1004-1021. Florin R 1938-1945. Die Koniferen des Oberkarbons und des unteren Pauze & Satter R 1978. L androcee centripete dochna atropurpurea. Perms, 1-8. Palaeontograpbtca 85B : 1-729. Canad.]. Bot. 56(20): 2500-2511. Rohweder 0 1967. Karpellbau und Synkarpie bei Ranunculaceen. Ber. Harris TM 1969. The Yorkshire]urasstc.flora. 3. Bennettitales. Brit. Mus. (Natur. Hist.), Lond. 186 pp. Schwetz. bot. Ges. 77: 376-425. Thomas HH 1925. The Caytoniales, a new group of angiospermous plants Heel WA van 1981. A SEM investigation of the development of free from the Jurassic rocks of Yorkshire. Phil. Trans. R. Soc. London carpels. Blumea 15 : 107-125. 8213 : 299-363. Herzfeld S 1922. Ephedra campylopoda Mey. Morphologie der weibli­ Tucker SC 1985. Initiation and development of inflorescence and flower chen Blute und Befruchtungsvorgang. Denkschr. Akad. Wiss. Wien in Anemopsiscal!fornica (Saururaceae). Amer.]. Bot. 72(1): 20-31. 98: 1-26. Vachrameev VA & Krassilov VA 1979. Reproductive structures of an­ Kaplan DR 1970. Comparative foliar histogenesis in Acorus calamus and giosperms from the Albian of Kazakhstan. Palaeont.]. 1 : 121-128 its bearing on the phyllode theory of monocotyledonous leaves. (in Russian). Amer.]. Bot. 57: 331-361. Walker JW, Brenner GJ & Walker AG 1983. Winteraceous pollen in the Khan R 1940. A note on double fertilization in Ephedra.foltata. Curr. Set. Lower Cretaceous of Israel: Early evidence of a magnoliacean an­ 9: 323-324. giosperm family. Science220(4603): 1273-1275. Krassilov VA 1969. Approach to the classification of Mesozoic Wettstein R 1907. Die Entwicklung der Blute der angiospermen Pflanzen ginkgoalean plants from Siberia. Palaeobotantst18: 12- 19. aus derjenigen der Gymnospermen. Wissen fur a/le 45 : 705-708