Terminal Inflorescence Research Articles

Morphology, biology, and systematic relationships ofDesmaria (Loranthaceae)

Desmaria mutabilis is unique inLoranthaceae in having dimorphic shoots, the short shoots producing a terminal inflorescence. Other unusual features in the family are well differentiated bud scales and deciduousness. The normal position of mature plants on the trunks of large trees is shown to be a consequence of profuse vegetative reproduction from the epicortical roots, the predominant growth direction of the latter towards the trunk from the original site of establishment on a lateral branch, and the ability of epicortical roots to generate haustorial contacts through heavy host bark. The seedling is heterocotylar, one cotyledon being phanerocotylar, the other cryptocotylar and functioning as a haustorial organ in the endosperm. It is suggested thatDesmaria is a member of the primitive complex of loranthaceous genera which includesGaiadendron.

Morphology and Pollination Biology of an Intersectional Hybrid of Costus (Costaceae)

An intersectional hybridization event between hummingbird-pollinated Costus pul- verulentus (sect. Ornithophilus) and large bee-pollinated C. guanaiensis var. macrostrobilus (sect. Costus) in Belize, Central America, is documented by analyses of morphology and pollination biology. Morphologically the hybrid is intermediate between the parental species and retains floral char- acteristics adaptive to both hummingbird and bee pollination. Nectar secretion rate and sugar concentration but not sucrose:hexose ratio in the hybrid are intermediate between those of the parental species. The hybridization event probably arose by breakdown of strict hummingbird pollination in Costus pulverulentus. Costus L. (Costaceae) occurs in moist habitats in the Neotropics, and is characterized by a spi- rally arranged phyllotaxy and showy terminal inflorescence. Costus is often found associated with species of Calathea, Canna, Heliconia, and Renealmia along the banks of rivers and streams and more disturbed habitats like forest trails, secondary forests, lightgaps, roadsides, and

Density and the commitment of apical meristems to clonal growth and reproduction in Hieracium pilosella.

We examined responses to population density in the commitment of apical meristems to reproduction and clonal growth in a rosette-forming, stoloniferous herb (Hieracium pilosella). Despite close physiological coupling between the evocation of the terminal inflorescence bud and the development of one or more axillary buds into stolons, the allocation of meristems was extremely plastic.Genets at the higher sowing densities showed density-dependent mortality consistent with self-thinning along a-3/2 trajectory. The probability of inflorescence evocation and associated stolon development was negatively dependent on surviving density. The proportinal distribution of primary stolons amongst genets became strikingly more unequal (expressed as the Gini coefficient) with increasing density. Clonal growth was resolved into the number of primary stolons per stoloniferous genet and the extent of stolon branching (i.e. number of apices per primary stolon); both showed strongly negative density-dependence. Reproduction, expressed as the mean number of flowering capitula per stoloniferous genet, declined 15-fold with increasing density; although theoretically expected to be unity, greater values resulted from capitulum production by attached secondary rosettes and lower values reflected the increasing abortion rate of inflorescence buds with increasing density.Both the total number of apices produced per unit area and the corresponding number of reproductive apices were maximal at intermediate surviving densities (700-1,000 m-2). The balance between reproductive and clonal growth may be expressed as the probability of an apical meristem producing a capitulum, that also peaked sharply at intermediate density. This finding does not conform with linear models that predict a shift from vegetative growth to sexual reproduction with increasing population density.

Systematic Anatomy of Cyclanthaceae (Monocotyledoneae)-An Overview

A recent survey of lamina and stem anatomy of Cyclanthaceae, including information from species that represent ca. 25% of the family, allows a comparison with Pandanaceae and Palmae, which have putative affinities with the Cyclanthaceae. The three families are readily distinguished by a number of discrete morphological and anatomical features. We present synopses and diagnostic keys with special emphasis on lamina anatomy. The subdivision of the Cyclanthaceae proposed by Harling is supported by anatomical evidence, but, because of anatomical similarities, related genera may not always be clearly separated solely by lamina anatomy. The Palmae are especially isolated. Their distinctive features include a closed tubular leaf base, silica bodies in specialized cells (stegmata), exclusively simple stem vascular bundles, and vessels (sometimes with simple perforation plates) in stem, leaf, and root. Pandanaceae are distinguished, among other features, by cubical crystal bodies, an open leaf base, compound stem bundles without basipetal fusion, and possible lack of vessels in stem and leaf. Cyclanthaceae have an open leaf sheath but lack both silica bodies and cubical crystals, and commonly have mucilage cavities and diverse structural features (laticifers, styloids, parenchyma-like dead cells, tannin-like star figures) that, however, are not ubiquitous. In stems, vessels with long scalariform perforation plates occur in Thoracocarpus. Compound stem bundles show basipetal fusion of a distinctive kind. Sympodial branching below a terminal inflorescence occurs in some Cyclanthaceae and especially in the Pandanaceae, but never in the Palmae. In contrast, features by which these three families are traditionally associated seem superficial.

Shoot ontogeny in Arctostaphylos uva-ursi (bearberry): the annual cycle of apical activity

Phenological investigation of shoot ontogeny in the prostrate shrub Arctostaphylos uva-ursi (L.) Spreng. (bearberry) at two sites in Saskatchewan, Canada, revealed that most growth occurred from May to July. Vegetative bud swell and leaf primordium initiation began around the 1st of May. Following bud burst in late May, elongation of most shoots continued for 3 to 5 weeks. Most bearberry shoots were not completely preformed; that is, several neoformed foliage leaves were initiated during current-year shoot extension in addition to the leaves that had been preformed during the previous season and had overwintered in the bud. In many shoots, a terminal inflorescence was initiated in the latter part of May of the year prior to anthesis. During conversion to the flowering state, the terminal apex initiated seven to nine floral bracts, each subtending a bud. In vegetative terminal shoots, bud-scale initiation also began in mid-May to late May and new terminal buds were first evident in early to mid-June. Following the initiation of bud scales and transitional leaves, the production of preformed foliage-leaf primordia continued until about August 1. Protruding lateral buds were evident histologically in the axils of preformed leaves during the initial stages of bud swell. On long, dominant shoots numerous neoformed leaves were initiated and shoot extension was often prolonged well into August. Second-flush terminal and lateral shoots, which resulted from the expansion of neoformed leaves and internodes, were also observed. The occurrence of neoformed growth in a large proportion of shoots suggests an exploitive, opportunistic growth strategy in this species.

Flowering-pattern and yield components at inflorescence nodes of snap bean as affected by irrigation and plant density

This paper reports the results of a 2 × 2 factorial experiment on bush snap beans ‘Oregon 1604’. The treatments were 2 contrasted irrigation regimes and 2 contrasted plant densities, and were applied in 1978 and repeated in 1979. Data were collected on the number of flowers and pods, and pod size, at each node of the terminal inflorescence (6-T) of the main stem, and at each node of the oldest inflorescence (2-A) at Node 2. High and low plant densities were 45 and 18 plants m −2 in 1978 and 54 and 33 plants m −2 in 1979. High temperatures, frequently above 32°C, prevailed during bloom and pod development in 1978, but for the most part occurred only during the week prior to bloom in 1979. Inflorescences 6-T and 2-A usually formed 4 and 3 RN's, respectively, in 1978 and 3 and 2 RN's in 1979. The flowers at the proximal nodes of each inflorescence all opened within a few days of one another (duration of flowering at proximal nodes between 3 and 5 days); the flowering-periods of adjacent nodes overlapped, and the flowering period increased acropetally within the inflorescence (duration of flowering at distal nodes between 7 and 13 days). In general, number of flowers, pods formed, pods harvested and percent set decreased acropetally within each inflorescence. The rate of acropetal decline was lessened by high irrigation or low plant density. In both years, high irrigation increased the percent set of all RN's of the 2-A inflorescences, but few other consistent effects between years were observed. The 2 most proximal RN's together produced 93% or more of the yield of each inflorescence. High irrigation significantly increased the total number of pods harvested from these RN's of inflorescences 6-T and 2-A, and low density had a similar effect on 2-A.

Effect of manual pinching, potting-on and cytokinins on branching and flowering of Camellia, Rhododendron and Rosa

Shoot pruning, or pinching, is commonly practised to promote branching and compactness in various nursery lines. In Camellia and Rhododendron pinching doubled the number of lateral shoots, but only if plants had been recently potted-on to encourage vegetative growth. Cytokinin (BAP) application had no effect on branching and appears not to be readily absorbed through the waxy leaves. By contrast, manual pinching reduced shoot number and growth in miniature rose, whereas BAP application promoted branching and growth. In addition, double the number of terminal inflorescences developed on the cytokinin-treated plants.

A Note on the (Presumed) Female Plant of Sphyranthera lutescens (Euphorbiaceae) in the Nicobar Islands

Sphyranthera lutescens (Kurz) Pax & Hoffm. is the sole known species of a monotypic genus established by J. D. Hooker in 1887 for a plant collected by S. Kurz on Middle Andaman Island 'in bamboo jungles along the Middle Straits'. Only the male plant was known; the female has remained unknown until the present day. Kurz referred his plant to the genus Codiaeum, but, as J. D. Hooker remarked when establishing for it the new genus Sphyranthera Hook.f. (S. capitellata Hook.f.), 'the affinities of this genus are uncertain; it is clearly not a Codiaeum, and is probably referable to the tribe Acalypheae' (tribe Crotoneae, subtribe Acalypheae, Hook.f., F.B.I. 5: 243). Codiaeum was included by Hooker in tribe Crotoneae, subtribe Chrozophoreae. These tentative expressions of opinion on taxonomic affinity were of course put forward on the basis of the male plant alone. Pax & Hoffmann, when dealing with Sphyranthera for the second edition of Engler's Pflanzenfamilien (1931) remarked: 'If belonging to the Euphorbiaceae probably to be inserted in the neighbourhood of Blachia' but I think this suggestion is negatived at least by the terminal inflorescence and thin leaves of Blachia. The genus was listed (with a description) at the end of the volume (19c: 231) among a number of other 'unsichere oder unvollkommen beschriebene Gattungen'; and so far as I am aware no other botanist has attempted to suggest the proper place for Sphyranthera within the Euphorbiaceae. In 1980 a collection was made from a fruiting tree in coastal forest at Laful on Great Nicobar Island (D. K. Hore 7579), not accompanied by corresponding male specimens, which I believe may represent the long-missing female plant of Sphyranthera lutescens. The duplicate at K consists of two leafy/fruiting scraps, bearing altogether five leaves, eight slender cauliflorous fruiting peduncles/pedicels, and a number of broken fragments (valves, seeds) of dehisced capsules in an attached envelope. The peduncles (0-5-1-5 cm) were evidently borne in sparse fascicles, were minutely fulvous-pilosulous or almost glabrous, and in turn bore a terminal few-flowered umbel of long-pedicelled flowers. The peduncles and pedicels of this collection (Hore 7579) are slender, firm and eventually almost wiry when dry, and persistent when the fruit has fallen. The capsules (apparently 3-locular) are externally smooth and glabrous, pale greyish in colour. The seeds appear to have been solitary in each loculus. Those that have survived are globose, with an obtuse inner angle, very smooth, of an ochraceous colour, about 4 mm diam., without aril, caruncle or other appendage. The 3 styles are short, deflexed, with numerous fine,

The delimitation and taxonomic position of the tropical African generaLeptactina andDictyandra (Rubiaceae)

The closely related African generaDictyandra andLeptactina are redelimited. The two genera are distinguished by long tubular corolla-tubes with included anthers (Leptactina) vs. much shorter tubes with exserted anthers (Dictyandra), rather than—as previously thought—presence or absence of multilocellate anthers. The discovery of a new species (Dictyandra congolana from the Congo and W-Zaire) confirms this. It can be concluded, especially from comparative morphological studies of fruits and seeds, thatDictyandra andLeptactina are related toPavetta, Tarenna, Ixora, etc., i.e. the group of genera in theCoffeeae s.l. with terminal inflorescences. It is proposed that the tribePavetteae be revived to accomodate those genera.

Supraspecific Groups in Tripsacum (Gramineae)

There are two main infrageneric groups within Tripsacum and they differ in several correlated characters. Variation is essentially continuous between the groups for any single char- acter but two well-defined modes of variation emerge when combinations of characters are consid- ered. There are few intermediates and the vast majority of specimens studied fall unambiguously into one group or the other. These two groups are recognized as sections. They are the most distinct and well-defined natural groups within the genus. Section Fasciculata differs from sect. Tripsacum in having one spikelet of a staminate pair at a rachis node with a long, thin pedicel rather than both spikelets subsessile or one supported by a short, thick pedicel; thinner internodes in the male section of each inflorescence branch; and a longer primary axis of the terminal inflorescence with more internodes and more branches. Section Tripsacum has a wide geographical range extending from the central and northeastern United States to central South America. Section Fasciculata has a more restricted distribution extending from southern Arizona to El Salvador. Tripsacum is a member of Andropogoneae, a large and morphologically well-defined tribe

Phenology and Control of Flowering in Tropical Trees

Descriptions of structure and flowering phenology of the tropical tree species Cedrela mexicana, Cordia glabra, Erythrina poeppigiana. Tabebuia rosea, and Tabebuia ochrea ssp. neochrysantha are used to analyze and illustrate the complex relations between structure, vegetative development, and flowering in tropical trees. Because of the strong positional differentiation within tree crowns and the repetition of vegetative and reproductive growth cycles in trees, flower development and its control are more complex than in herbaceous plants. Formation of terminal inflorescences, as found in Cedrela, Cordia, and Tabebuja, regularly occurs at the end of a flush of determinate shoot growth; flower initation, like shoot growth cessation, is therefore controlled by internal, correlative factors rather than by environmental changes. Initiation of lateral inflorescences is usually preceded by growth cessation in the indeterminate, vegetative shoot, but the timing and control of flower induction in lateral buds are unknown. For many tropical trees, particularly those growing in seasonal climates, flower development is discontinuous, i.e., flower initiation and anthesis are separated by a prolonged rest period and must be controlled separately. Delayed anthesis is usually triggered by rehydration of previously water-stressed trees resulting from leaf shedding or, in dry habitats, by rainfall. Because of different degrees of seasonal water stress, leaf fall and subsequent anthesis in several species extend over three to four months. The observed periodicity of anthesis in tropical trees is thus largely determined by seasonal changes in tree water status and is unlikely to be the result of selection for optimum tree-pollinator interaction. IN MOST TROPICAL TREES, flowering, like vegetative growth is episodic, and seasonal peaks of flowering have been recorded for many tropical forests (e.g., Fournier and Salas 1966, Janzen 1967). During recent years, tropical ecologists have considered the temporal pattern of flowering the result of coevolution of trees with pollinators and seed predators. It has been argued that peaks of flowering occur during periods of high pollinator activity, and that overlapping, but different, flowering periods in related species reduce competition for pollinators and assure a continuous source of nectar to pollinators (Janzen 1967, Gentry 1976, Stiles 1978). If the temporal pattern of flowering is adaptive, then mechanisms for the specific timing of flowering must have evolved in tropical trees. The timing of flowering in tropical trees usually has been ascribed to environmental control mechanisms analogous to those that have evolved in temperate plants to adapt vegetative and reproductive development to a growing season of limited duration (Larcher 1980). Relations between tree architecture, vegetative growth and flowering of tropical trees have been rarely considered. In this paper, I shall use descriptions of the flowering behavior of five tropical tree species as a basis for discussing the relations between tree architecture, vegetative development, and flowering, as well as our scant knowledge of their internal and environmental controls. I will show that evolution of the temporal pattern of flowering in tropical trees cannot be adequately explained as resulting from biotic interactions alone, but must be seen in the context of overall tree development as determined by a given set of climatic, edaphic, and biotic factors. STUDY SITES AND METHODS Most of the phenological observations on flowering of tropical trees discussed in this paper were made on the Pacific slope of Costa Rica, Central America, along an altitudinal gradient extending from the deciduous lowland forest of Guanacaste Province (Janzen 1967, Daubenmire 1972, Frankie et al. 1974, Opler et al. 1976, Reich and Borchert 1982, 1984) through the lower montane forest at Villa Col6n (800 m altitude) and Santa Ana (900 m; Fournier and Salas 1966, Fournier 1969) to the Central Valley of Costa Rica (La Sabana, 1100 m; San Pedro Montes de Oca, 1200 m; Granadilla, 1350 m) (Fournier 1969, Borchert 1980). The climate in the entire area is characterized by a dry season lasting from late November to April/May and a wet season providing adequate rainfall (approx. 2000 mm) for the remainder of the year. With increasing altitude, the severity of the dry season diminishes and some species change from a deciduous to an evergreen habit (e.g., Erythrina poeppigiana, Tabebuia rosea; Fournier 1969, Borchert 1980). The discussion of tree morphology is based on the analysis of tropical tree architecture by Hall6 et al. 1978, where definitions of all morphological terms can be found. MORPHOLOGY AND PHENOLOGY OF FLOWERING IN FIVE TROPICAL TREE SPECIES TABEBUIA ROSEA (BERTOL.) DC, BIGNONIACEAE.-The shoot system of T. rosea is composed of many orthotropic, symBIOTROPICA 15(2): 81-89 1983 81 This content downloaded from 157.55.39.17 on Fri, 02 Sep 2016 06:06:02 UTC All use subject to http://about.jstor.org/terms

Food-Foraging Behavior of Male Euglossini (Hymenoptera: Apidae): Vagabonds or Trapliners?

Food-Foraging Behavior of Male Euglossini (Hymenoptera: Apidae): Vagabonds or Trapliners?

Bird Visitation, Seed-Set, and Germination Rates in Two Species of Lobelia on Mount Kenya

Seed-set in natural plant populations may be limited by pollinators, resources, or seed predators. Bierzychudek (1981) found pollinator numerical limitation of seed-set to be common in natural populations. However, many plant species are capable of self-pollination when pollinators are absent (Proctor and Yeo 1973); in others natural levels of outcrossing are sufficient for full seed-set, and seed-set appears to be resource limited (Primack and Lloyd 1980, Udovic 1981, Udovic and Aker 198i). In most pollination studies of natural plant populations, numerical seed-set is the only measure of reproductive success, although recent studies have shown that other factors such as seed size, germination rates, and seedling success may be affected by the degree of outcrossing (Price and Waser 1979, Schoen 1982). Such effects have long been known for crop plants (Crumpacker 1967). This paper reports an attempt to determine the importance of pollinators for the seed production of two species of Lobelia on Mount Kenya. Two species of rosette plants in the genus Lobelia (family Lobeliaceae) occur in the alpine zone of Mount Kenya. L. telekii is primarily semelparous, and L. keniensis is primarily iteroparous. These two species are thought to be bird pollinated (Hedberg 1964). The morphology of the reproductive structures is such that birds, but no native insects, normally come into contact with the pollen source (T. P. Young, personal observation). The cylindrical inflorescences project above the surrounding vegetation, and the purple flowers secrete copious nectar. L. telekii has narrow, hairy bracts and is visited primarily by territorial individuals of the Scarlet-tufted Malachite Sunbird (Nectarinia johnstonii). L. keniensis has broad glabrous bracts, and is visited by a variety of birds, including N. johnstonii, the Mountain Chat (Pinarchroa sordida), and the Slender-billed Chestnut-winged Starling (Onagnathus tenuirostris). The tall terminal inflorescences mature progressively upwards from the base. The five petals are united and open between the upper two petals to form a lower-lipped flower (Fig. 1). The style is enclosed by a tube formed of the five united filaments and anthers. This tube terminates in a brush that acts as a block to pollen release. Dehisced pollen collects inside the distal end of the filament tube. When the brush is pushed back by the forehead of a visiting bird, pollen is released. I have observed white pollen on the foreheads of P. sordida and 0. tenuirostris visiting L. keniensis flowers. The male state of the flowers lasts up to 2 wk. The style lengthens during this time, and is finally exserted from the filament tube. The stigmas separate and become receptive several days later, and begin to dry up after 2-4 wk. Seeds develop in a pair of inferior ovaries and are wind dispersed over a period of many weeks. The time from the first appearance of the flowers to final dispersal of seeds is =12-18 mo, depending

THE BALANCE BETWEEN CHASMOGAMY AND CLEISTOGAMY IN COLLOMIA GRANDIFLORA (POLEMONIACEAE)

The extent of cleistogamy (CL) and chasmogamy (CH) was investigated in natural populations of Collomia grandiflora Dougl. ex Lindl. The ubiquitous occurrence of CL flowers was documented in 18 populations from throughout the geographic range of the species. The interrelationship of environmental perturbation, vegetative weight, and the balance between CH and CL flowers was investigated by means of a factorial experiment which varied plant density, incident light and soil mix. An ANOVA revealed that total plant weight increased significantly under conditions of low density, low light and low sand content, with density accounting for most of the variance in plant weight. Interactions involving density were not significant. The number of CH and CL flowers and estimated percent outcrossing increased significantly with increased plant weight, low density, low light, and low sand content, although an analysis of covariance showed that plant weight and density accounted for most of the variance in these traits. Among large, multibranched plants, the proportion of CL flowers increased in a basipetal pattern, whereas most of the CH flowers were restricted to the terminal and upper inflorescences. The smallest plants produced only a single terminal inflorescence comprising entirely CL flowers. The proportion of reproductive effort allocated to outcrossing was dependent upon plant size as estimated from total weight, whereas the proportion allocated to ***selfing was relatively independent. Studies of outcrossed vs. inbred progenies of five populations revealed little evidence for immediate inbreeding depression with respect to plant weight and the number of CH or CL flowers, but indicated that significant quantitative genetic differences exist for these traits among populations of diverse ecological or geographical origin.

Developmental Potential of Axillary Buds of Water Hyacinth, Eichhornia crassipes Solms. (Pontederiaceae)

Buds axillary to foliage leaves of water hyacinth can elongate either as vegetative stolons or as renewal shoots produced in association with the terminal inflorescence. Stolons differ from renewal shoots in position within the shoot system, morphology, and function. Renewal shoot buds always expand, whereas stolon buds may or may not. A stolon bud develops in conjunction with the subtending leaf; as that leaf matures, the stolon bud reaches a critical period in development. At this point, the bud either continues to expand, producing a stolon, or it stops growth and matures. Maturation is not irreversible, but the probability of a bud expanding decreases as bud age increases. In the field, buds on plants at the water hyacinth mat edge frequently produce stolons, whereas buds on plants inside the mat rarely do so. Leaf morphology also varies between plants in these two regions of the mat. The particular association of leaf and branch type found in the field, however, can be reversed experimentally, indicating that although leaf and bud development are coordinated, the particular course of each is independent.

DEVELOPMENTAL POTENTIAL OF AXILLARY BUDS OF WATER HYACINTH, EICHHORNIA CRASSIPES SOLMS. (PONTEDERIACEAE)

Buds axillary to foliage leaves of water hyacinth can elongate either as vegetative stolons or as renewal shoots produced in association with the terminal inflorescence. Stolons differ from renewal shoots in position within the shoot system, morphology, and function. Renewal shoot buds always expand, whereas stolon buds may or may not. A stolon bud develops in conjunction with the subtending leaf; as that leaf matures, the stolon bud reaches a critical period in development. At this point, the bud either continues to expand, producing a stolon, or it stops growth and matures. Maturation is not irreversible, but the probability of a bud expanding decreases as bud age increases. In the field, buds on plants at the water hyacinth mat edge frequently produce stolons, whereas buds on plants inside the mat rarely do so. Leaf morphology also varies between plants in these two regions of the mat. The particular association of leaf and branch type found in the field, however, can be reversed experimentally, indicating that although leaf and bud development are coordinated, the particular course of each is independent.

Influence of 6-Benzylaminopurine on Fruit-Set and Seed Development in Two Soybean, Glycine max (L.) Merr. Genotypes

The influence of 6-benzylaminopurine (BA) on the premature abscission of developing soybean, Glycine max (L.) Merr. fruits of 2 genotypes was studied. BA was applied during the critical period of fruit-setting. The tested concentration range of BA was from 1 micromolar to 5 millimolar; 2 millimolar was optimal. Spray application of 2 millimolar BA to terminal inflorescences at the R(3) developmental stage of field-grown soybeans significantly increased fruit-set and seed yield of the Shore genotype during three growing seasons. In contrast, the Essex genotype gave significant responses two out of three seasons. The response of Shore was generally more pronounced than that of Essex. The apical fruits on the inflorescences gave the greatest response to BA. Seed weight increase was apparent 3-4 weeks after BA treatment.

Interactions of the Growth Retardants Daminozide and Piproctanyl Bromide, and Gibberellins A1, A3, A4+7, A5 and A13 in Stem Extension and Inflorescence Development in Chrysanthemum morifolium Ramat

Interactions between the growth retardants daminozide (a substituted succinamic acid) or piproctanyl bromide (a quaternary ammonium, piperidinium compound), and a subsequent application of a single dose (40 μg) of either gibberellin A1, A3, A4+7 or A12, showed that, in Chrysanthemum morifolium Ramat. cv. Bright Golden Anne, a strong interdependence exists between elongation of the lateral shoot and the rate of development of its terminal inflorescence. A1, A3, and A4+7 were highly active in overcoming the restrictions on both internode extension and the rate of flower-bud development imposed by either retardant, suggesting that these two retardant effects are caused by a deficiency of active gibberellins (GN). In the absence of retardant, A1, A3, and A4+7 markedly increased stem elongation, and flowering occurred earlier than in plants receiving neither retardant nor GN. A13 the only 20-carbon GN tested, was much less active, while A5 had a relatively greater effect on the time of flowering than on shoot elongation. Thus, it is not necessarily the rate of stem extension which determines the rate of inflorescence development. The response to different amounts of A1, A3 or A13 (1, 5, 10, 20, or 50 μg per shoot) neither suggest that different ‘threshold’ levels of a particular GN are required to induce increases in either stem elongation or in the rate at which inflorescences develop, nor did a change in the dose given lead to any consistent differential effect on these two processes.

Phenology of reproductive shoots of eelgrass, Zostera marina L., at Roscoff (France)

The development of flowering shoots and the maturation of flowers and fruits of Zostera marina L. were studied from May until the beginning of September 1978 at Roscoff (France) at two different tidal depths. Branches of flowering shoots matured acropetally, with the youngest branch as terminal inflorescence, and this also applied to the spathes within each branch. So, the development and maturation appeared to be a process starting at the base and proceeding to the apex of the shoot. Within spathes, the number of flowers varied greatly. A close examination of the first order spathes showed that the number of flowers per spathe was positively correlated with increasing tidal depth and branch number and negatively correlated with the initiation time of the spathes. However, the ratio of one pistil to one anther was always found, although at the apex of the spadix some anthers sometimes failed to mature. A direct relation between the number of flowers per spathe and the mean length of the spadix and the spatial sheath could not significantly be demonstrated. The mass flower production had already begun in the high-littoral bed in May, whereas this occureed in the lower bed one month later. However, the development and maturation of flowers and fruits occurred more quickly in the low-littoral bed than in the higher bed. As the ripening of fruits did not occur simultaneously, even within one spathe, this process could not be followed in situ.

Cuphea inflata (Lythraceae), a New Species from Mexico

Cuphea inflata S. Graham is described as a new species of sect. Heterodon from Nayarit, Jalisco, and Sinaloa. It is distinguished from its closest relative, C. leptopoda Hemsl., by its secund or pyramidal inflorescence, wine-purple floral tube inflated at maturity, and chromosome number, n = 10. Recognition of a new species of Cuphea from Nayarit, Mexico, is supported by new morphological and cytological data gathered in recent field work. Earlier collections of the species were determined as C. leptopoda Hemsl. or as having affinities with that species. Cuphea inflata S. Graham, sp. nov., figures 1-4. Herba annua; folia anguste elliptica ad elliptica; folia floralia abrupte minora, bracteiformia. Inflorescentia terminalia, secundiflora vel pyramidalis, panicularis 1-multiflores composita. Calyx 12-17 mm longus, dorso manifeste vinosus in statu vivo, atro-purpureus in statu sicco; calyx fructifer ampulliformis, inflatus infra medium, pallescens. Petala 6, duo dorsalia vinosa, 4 ventralia lilacina. Ovula 3. n = 10. TYPE: Mexico, Nayarit, 3 km E of Santa Maria del Oro on road to the lake, oak zone, open sunny areas along roadsides, 20 Aug 1978, S. Graham 608 (holotype: MICH; isotypes: GOET, MEXU). Herbaceous annual, erect, 20-75 cm tall, sparsely branched below, the branches shorter than the main stem; stems green, very viscid, bearing dense, glandular, purple hairs in the inflorescence. Petioles 10 (5-20) mm long; leaf blades narrowly elliptical to elliptical, 30-65 mm long, 1230 mm wide, minutely uniformly scabrous on upper and lower surfaces, the base acute, the apex acuminate;' uppermost leaves abruptly reduced in size becoming linear, ciliated bracts of the terminal inflorescence. Inflorescence compound, shortto long-paniculate, secund or pyramidal on mature plants, the branches of the panicle becoming progressively shorter and fewer-flowered toward the apex of the plant, the longest branches up to 8 cm long with ca. 20 flowers, most branches 2-3 cm long with ca. 5 flowers; flowers 1-5 per node; pedicels 1-2 mm long, bibracteolate, the bracteoles composed of 1-3 hairs, scarcely developed. Calyx 12-17 mm long, distinctly wine-colored, drying purple-black, the ribs dark purple bearing purple glandular hairs, the base rounding, barely spurred, the lower distal end saccate, the calyx ampullaceous at maturity, the proximal end becoming exceptionally inflated and pale pink in con-