Abstract: IT has been known since ancient times that nectar, usually secreted from specialized glands within flowers, is utilized by anthophilous (flower-visiting) animals for the energy-providing sugars that it contains. It is usually assumed that protein-making materials will be obtained elsewhere, from larval feeding (Lepidoptera), from pollen (Hymenoptera—Apioideae) or (by vertebrates) from insects. However, some butterflies are known to be attracted to decaying flesh, faeces, urine and stagnant water1 and well as sweat2, while many take phloem-sap from a wound or honey-dew excreted by aphids1. Moths are known to drink fruit juices, sweat, secretions from the eyes of animals, and even mammalian blood3. Unusual feeding-behaviour is shown by butterflies of the genus Heliconius which collect pollen, steep it in nectar, and subsequently ingest the amino-acids that diffuse from the grains4.

Abstract: Evidence from sediment traps shows that in dimictic Frains Lake pollen grains and the bottom mud in which they are contained are rcsuspendcd during spring and autumn mixing, while in Saylcs Lake, a nonstratificd lake nearby, similar quantities of pollen-bearing sediment are resuspended at irregular intervals throughout the year. In the nonstratified lake resuspended material is poorly mixed in the lake water, while in the dimictic lake, during the seasons of water mixing, resuspended sediment is similar in amount, perccntagc a5 .h weight, and pollen composition at various lcvcls throughout the water column. Rcsuspension occurs without sorting or diffcrcntial movement of individual pollen grains. The pollen content of rcdepositcd scdimcnt serves as a tracer, showing that sediment is moved from the littoral zone to the deeper basin of the lake. In the littoral zone annual stirring may involve the uppermost 6-12 mm of sediment; even in the dccpcr part of the basin, the uppermost millimeter at least is stirred by this process every year. Redeposition of sediment and pollen has been rcportcd from experiments with scdimcnt traps in Frains Lake, a thermally stratified lake in Michigan (Davis 1968); circulating water stirs and rcsuspends scdimcnt during the spring and fall seasons of water mixing. This process is capable of redistributing scdimcnt within the lake basin (Tutin 1955) and thus affecting the Final distribution of pollen grains in sediment, so it has obvious importance for the interpretation of fossil pollen. Seasonal measurcmcnt of redcposi tion has been cxtcnded to a shallow, nonstratified lake for comparison with Frains Lake. The constrast in timing demonstrates that absence of thermal stratification is csscntial for extcnsivc rcdcposition. To evaluate movcmcnt of the scdimcnt and its effect on pollen distribution, scdimcnt traps wcrc set at various depths in various parts of both lakes. With pollen grains serving as a tracer, thcsc traps dcmonstratc the movcmcnt of scdimcnt from one part of the lake basin to another. They serve further to detect any sorting due to differential movement that may occur within the pollen component of the scdimcnt, a phenomenon that had been suspected from the uneven disL Contribution No. 147 from the Great Lakes Restarch Division, The University of Michigan. tribution of different kinds of pollen within lake basins (Davis et al. 1971). The total amount of pollen deposited in traps at diffcrcnt stations scrvcs to assess the thoroughness of water mixing, In addition to information on the dircction of sediment movement, the amount of movcmcnt, and the distribution of pollen within the suspcndcd sediment, the results of the study provide an estimate of the depth to which sediment in lakes is stirred by the action of water currents. Throughout this paper rcferencc is made to “pollen grains.” In fact the refcrcncc is to the waxy exine of the grains, which is identifiable taxonomically and which is rcsistant to decay. Pollen grains filtered from lake water during the flowering season still retain protoplasm, but a few weeks or months later, only the exine remains. Ccrtdn cxccptional types of pollen arc complctcly destroyed soon after entering the lake by bacteria or fungi (Havinga 1967) or benthic organisms (R. Davis 1969). For example poplar pollen was foulld in traps during the flowering season, but was virtually absent from the sediment on the lake floor. However, pollen is not destroyed after it has been incorporated in sediment; this is cstablishcd by the constancy through time of the ratio of pollen grains to silt and clay in the sediment (Davis 1968). It is for LTMNOLOGY AND OCEANOGRAPIIY 44 JANUARY 1973, V. 18( 1) REDEPOSITION OF POLLEN GRAINS 45 this reason that pollen serves as a useful tracer for suspension, movement, and rcdeposition of the scdimcnt as a whole. I gratefully acknowlcdgc the assistance of J, IM. Bciswengcr and L. B. Brubakcr in the design and cxccution of thcsc cxpcrimcnts. W, R. Solomon kindly made available his data on pollen content in the air at Ann Arbor. The George Rcscrvc, The University of Michigan, pcrmittcd the USC of sediment traps on Sayles Lake in 19651966. The work was supported by National Science Foundation Grants G,B 2377, GB 5320, and GB 7727. DESCRIPTION OF SITES Frains Lake (42”20’N, 83”37’W), 12 km northeast of Ann Arbor, Michigan, is about 200 m wide and 500 m long, sloping gradually to a maximum depth of 10 m in the center. There arc no inflowing or outflowing streams. The surrounding topography is nearly flat, (2uitc unprotcctcd from wind, the lake is surrounded by meadows with a relief of only 3-5 m. ICC about 30 cm thick persists for 3-4 months each winter. In summer the water is strongly stratified, with the top of the thermocline at about 4 m. (For more dctailed information on topography, see Davis ct al. 1971.) Saylcs Lake (42”26’N, 84”4’W), near Pinckncy, Michigan, and 35 km west of Frains Lake, is similar in area and shape to Frains Lake and, like Frains Lake, has no major inflowing or outflowing s trcams. Howcvcr, the lakes do differ in two major ways: 1) Saylcs Lake is less than 1 m deep over most of its area and only rcachcs a maximum depth of 3 m in an area just south of the basin’s ccntcr, and 2) Sayles Lake is surrounded by woods on three sides with a steep ridge 20 m high along the south short. ICC cover is similar to Frains Lake, but Saylcs Lake ncvcr develops thermal stratification.

Abstract: Anthers of Triticum aestivum L. were cultured in vitro on the MS medium supplemented with auxins, with or without kinetin. Then the pollen grains were successfully, induced to develop into intact plants through a stage of callus formation. Experiments indicated that pollen grains at the mid-uninucleate stage were most favourable for induction to form calli, and that addition of lactalbumin hydrolysate and appropriate increase of the sucrose coneentration had some promoting effects on pollen callus formation. The pollen callus with relatively great differentiating capacity was generally compact in texture and possessed the histological characteristics of the primary meristem. It was noted that the results obtained with different materials (different varieties or hybrids of different crosses) varied considerably. The progeny of the plant induced from F 1, hybrid pollen showed no segregation of characters, suggesting that application of the anther culture method to plant breeding is very promising.

Abstract: The tryphine that coats the pollen grains of Raphanus is tapetally synthesized and is composed of a fibro-granular and a lipidic component. The fibro-granular material is proteinaceous and is secreted by cisternae of the endoplasmic reticulum. The lipidic component is derived, mainly, from degraded elaioplasts. The fibro-granular material is applied to the pollen exine first, followed by the lipidic mass. The tryphine condenses during the final stages of pollen maturation and dries down to form a thick, highly viscous coating. The major part of the condensation appears to result from dehydration. The tryphine, extracted from the pollen by a centrifugal method and mounted in a membrane, appears to be capable of penetrating the outer layers of a stigma of the same species and, if the pollen from which it was derived is incompatible with respect to the stigma, the stimulation of the production of the callosic reaction body in a manner similar to an incompatible pollen tube. It is proposed that, in Raphanus, substances responsible for the initiation of at least two stages in the self-incompatibility system are held in the tryphine.

Abstract: Detached pistils of Lilium longiflorum were labeled with d-glucose-U-(14)C 24 hours after anthesis and then sampled for the next 6 days to determine the appearance of label into exudate from the stylar canal and the stigmatic surface of the pistil. Results were obtained with unpollinated cv. Ace pistils and pollinated cv. Ace pistils, selfed or crossed (cv. Croft pollen). Limited data were also obtained on cv. Croft pistils, selfed or crossed (cv. Ace pollen).Exudate appeared in the canal and on the stigmatic surface soon after anthesis. In unpollinated pistils it continued to accumulate for about 5 days in the canal and for the full term of the experiment, 7 days, on the stigmatic surface. Canal exudate eventually mixed with stigmatic exudate in unpollinated pistils. Pollination interrupted the flow of exudate, and a portion of the pistil secretion product was diverted toward the synthetic requirements of the developing pollen tubes. Two days after pollination, the specific radioactivity of pollen tube cytoplasm had reached a level comparable to canal exudate. Twenty-four hours later, the specific radioactivity of pollen tube wall substance was about 80% of the value found in canal exudate. Similar patterns of (14)C incorporation and similar carbohydrate contents were obtained from self- and cross-pollinated pistils, although the former contained pollen tubes of less than one-half the length of the latter.

Abstract: Pollen grains from 150 Nigerian woody plants, comprising over 60% of the recognised climatic indicator species of the forest and savanna zones, are described, many of them for the first time. 73 are also illustrated. Some of the observations made are compared with those of earlier investigators. Intra-specific variations were noted.

Abstract: Embryoid formation was induced in pollen of Solanum tuberosum, cv. Pentland Crown, Maris Piper and Record, by culturing anthers at or just prior to the first pollen mitosis. The composition of the medium was not critical in respect of either nutrients or hormones. In the presence of relatively high hormone levels, callus formation also occurred from somatic tissues of the anthers.

Abstract: Palynological investigations of spring deposits were conducted as part of an interdisciplinary research effort into late Pleistocene archaeology, geology, and paleoecology of the western Missouri Ozarks. The springs, located in the Pomme de Terre River valley in Benton and Hickory Counties, Missouri, lie 400 km south of the maximum extent of the Wisconsin ice sheet on the present Great Plains—Eastern Deciduous Forest border. The valley will soon be flooded by a government impoundment. Two springs were completely excavated and three others sampled, including a spring originally excavated by Albert Koch in 1840. The pollen record, with associated plant macrofossils, is divided into three distinct zones: a lower nonarboreal pollen (NAP) and pine zone, a middle spruce— (Picea) dominated zone, and an upper zone of spruce with deciduous elements. Radiocarbon dates show that the NAP—pine zone was established prior to 40,000 BP and existed until 20,000—25,000 BP; it was deposited during the mid—Wisconsin interstade. The mid—Wisconsin interstadial vegetation of the Ozarks is interpreted as an open pine—parkland. Plant macrofossils of Pinus banksiana in this zone indicate that jack pine was present. Spruce was probably absent as no spruce macrofossils and only traces of spruce pollen were found. The associated mid—Wisconsin fauna is dominated by mastodon (Mammut americanum) with horse (Equus) and musk—ox (Symbos). With the onset of late Wisconsin full—glacial conditions, variously dated at 20,000—25,000 BP, pollen dominance shifts from NAP—pine to spruce. This is interpreted as the replacement of the pine—parkland by boreal spruce forest. Two pollen zones are apparent within the full—glacial: an older zone with up to 92% spruce pollen occurring after the transition from NAP—pine to spruce, and a younger zone with less spruce (38%) and more thermophilous deciduous tree pollen. The change to less spruce and more deciduous tree pollen apparently reflects slightly improved climatic conditions during a late phase of the full—glacial; the zone is radiocarbon dated at 16,500 BP on spruce logs. It contains abundant spruce and larch (Larix) macrofossils associated with the remains of at least 30 mastodons plus giant beaver (Castoroides), ground sloth (Paramylodon), tapir (Tapirus), deer (Odocoileus), and horse (Equus). The western Missouri Ozarks were covered by open pine—parkland from at least 40,000 BP until the start of the full—glacial 20,000—25,000 BP, then boreal spruce forest until at least 13,500 BP. Deciduous elements became more prominent in the spruce forest in its latter phase. The present oak—hickory forest developed after the decline of the spruce and is thus no older than postglacial.

Abstract: The female and male flowering frequency were studied in a clone trial of Picea abies at Roskar nine kilometers north-east of Stockholm During 1971 the extension in time of pollen shedding and female receptivity was recorded following daily examination of individual strobili The data obtained revealed a great variation in female as well as male flowering between the clones Only a part of this variation could be attributed to differences in height of the clones Great yearly variations in flowering of individual grafts were noticed Based on the flowering frequency as well as the pollen shedding and female receptivity during each day, the expected contribution of the different clones to the offspring was calculated According to the calculations four of the clones contributed 55 per cent of the genes to the offspring Many (66) of the 190 theoretically possible combinations among the 20 clones occurred in a lower frequency than 01 per cent The consequences of the data obtained for the genetic composition of the seed orchard progeny were discussed

Abstract: The Chuska Mountains, with an elevation of 2,700 m above sea level, are covered with a forest of ponderosa pine punctuated by openings around small lakes. Spruce, fir, Douglas-fir, and aspen are found on the north-facing slopes of canyons. At lower elevations on the mountain flanks, Gambel oak accompanies the ponderosa, and this altitudinal belt gives way below to a woodland or savanna of pinyon pine and juniper, with sagebrush between the trees. The San Juan Basin east of the mountains is dominated by shrub steppe, with various chenopods and composites. Plots of herbs and shrubs in the various forest belts show an equally clear altitudinal gradient in occurrence and abundance. The vegetation history is worked out by pollen analysis of sediments of four lakes from the mountain crest. An 11-m core from Dead Man Lake gives the longest record, representing perhaps the last 50,000 yrs. Pollen zone 1, of Holocene age, is only a few decimeters thick. It is dominated by pine pollen, derived from vegetation similar to today9s for the area. Pollen zones 2 to 5 are dominated by Artemisia pollen and further characterized by spruce as well as pine pollen. Separation of pine pollen types on the basis of size and shape indicates the past presence of Pinus flexilis (limber pine) in the area, even though this tree is not present there today. This assemblage is interpreted as a record of alpine vegetation of Pleistocene age, with the tree pollen having been blown up to the crest from below. The upper forest limit is believed to have been lowered about 900 m for zones 2 and 4, and even more for zones 3 and 5, so that the upper parts of the mountain flanks were covered with spruce/fir forest. Successively lower forest belts were not depressed so much, however: ponderosa pine did not reach the base of the mountains but was restricted to the lower flanks, and pinyon pine and Artemisia spread across the San Juan Basin. This telescoping of altitudinal belts implies a climatic gradient steeper than that of today.

Abstract: The release of wall-held materials from the pollen of ragweed ( Ambrosia tenuifolia ) and Cosmos bipinnatus on to the stigma surface has been followed. When fresh stigmas were viewed by scanning electron microscopy, a fluid material was observed coating pollen grains, pollen tubes and adjacent stigmatic papillae. This fluid contained pollen-wall antigens, including the allergen Antigen E, detected by immunofluorescence, and proteins, lipids and carbohydrates detected by cytochemical methods. In Cosmos , the fate of the antigens was much the same after both compatible and incompatible matings. In incompatible matings, pollen tubes were blocked with a polysaccharide considered to be callose. Callose particles appeared on neighbouring stigmatic papillae, and adjacent stigmatic hairs accumulated callose internally. This reaction may be an important one for the incompatibility response.

Abstract: In many orchards the crop of fruit is reduced in some years by inadequate pollination (Williams & Wilson 1970). The chief agents of pollen transference are honeybees, the numbers of which the grower may augment by keeping or hiring colonies. Wild pollinating insects are also important, especially when, as is often the case, the numbers of honeybees are inadequate. It is desirable to know something of the relative merits of different species as pollinators of fruit blossom for the choice of candidates for protection and encouragement, or simply to assess their present contribution. Various factors play a part in determining how effective a particular species is likely to be as a pollinator (Free 1970; Smith 1970), including its frequency of movement from flower to flower and from tree to tree, and the extent to which it brushes against the anthers and stigmata when visiting a flower and feeding on pollen or nectar. The amount of pollen adhering to the body is obviously important in this connection, and this varies between species. Most small insects with smooth bodies carry little or no pollen, whereas large 'hairy' insects bear considerable amounts. In this paper we are concerned only with the amounts of pollen on the insects' bodies. A complicating factor is that the foraging castes of the non-parasitic species of bees collect pollen into 'pollen baskets' on the hind legs. Honeybees and bumblebees moisten this load with nectar or honey, and normally it adheres well during foraging, and if dislodged it falls away en masse. This is less true of the pollen collected and carried in a dry mass on the hind legs of solitary bees. Nevertheless, as with honeybees and bumblebees, we have assumed that this pollen is not of much significance in pollination and have always removed the hind legs when they bore such a load before separating and counting the pollen dispersed over the rest of the body. So far as we know, the only other work along these lines is by Lukoschus (1957) and Skrebtsova (1957).* They estimated the amounts of pollen on the bodies of honeybees, using different methods of separating the pollen from that we adopted. Some of the honeybees examined by Lukoschus were from apple blossom. He examined some other insects, chiefly flies, but all were taken from Solidago canadensis L., which differs considerably from apple in its pollen and its flowering season (late summer). Skrebtsova (1957) examined honeybees frequenting the flowers of various crop plants, including a few from the blossom of certain fruits, and also estimated the amounts of pollen on different parts of their bodies.

Abstract: Between 1957 and 1966, pollen shedding by longleaf pine (Pinus palustris Mill.) in southwestern Alabama peaked at dates ranging from February 23 to April 3. January 1 and 50°F was the combination of starting date and threshold air temperature that minimized annual variations in heat sums before the trees flowered. The heat sum required for peak pollen shedding declined as the season advanced. The regression Y = 19009 -89.26X, in which Y is the degree-hour heat sum above 50°F and X is days from January 1 through date of peak pollen shedding, accounted for nearly all observed annual variations in peak date. For the data from which the regression was computed, deviation of observed from expected date of peak pollen shed averaged 0.3 day. For eight subsequent observations, including two in North Carolina and one in northern Alabama, the deviation averaged 1.6 days. Flowering by the four major southern pines fol- lows a consistent pattern (Dorman and Barber 1956)) with slash pine (Pinus elliottii Engelm.) first, fol- lowed by longleaf (P. pdustris Mill.), loblolly (P. taedu L.), and shortleaf pine (P. echinata Mill.). In the Gulf States (Wakeley 1954), slash pines normally shed their pollen in late January or early February, longleaf and lo~blolly pines in March, and shortleaf in April. A nugmber of investigators have concluded that air temperature during the days or weeks before flowering is the major factor governing the time of pollen shedding by pines (Fielding 1957, Marcet 19.51, Millett 1944, Minshall 1947, Scamoni 1938, Snyder 1961). Pollen shedding by longleaf pines has been ob- served for 15 years on the Esoambia Experimental Forest in the Coastal Plains of southwestern Al- abama, 2 years in jthe fall-line sandhills of North Carolina, and 1 year in the Mountain Province of northern Alabama. Southwestern Alabama data for the 10 years from 1957 through 1966 were analyzed to explore the relationship between air temperatures and longleaf flowering, and to determine the degree to which this factor may account for annual vari- ations in the date of peak pollen shedding. In studies of heat effects upon plant development, heat is usually expressed in degree-hours or degree- days above a biologically established threshold tem- perature. The heat unit system was developed in the eighteenth century, and most subsequent development and applications of the method have been in agricul- ture (Holmes and Robertson 1959). In most re- search, the heat sum required to attain a given stage of development has been assumed to be a fixed value, the "thermal constant," although actual constants have rarely been found in phenological studies. Heat sums associated with a phenological event for a given species normally ,vary from place to place,

Abstract: Summary A general survey of the pollen-rain from the vegetation of north-west India is described. The survey is based on the study of 114 modern pollen spectra from sixty-four different sites. It is revealed that with certain qualifications the major aspects of the extant vegetation belts of north-west India are reflected in the pollen-rain falling within them. In the territory as a whole, the pollen of herbs dominates that of trees and shrubs except in close proximity to one of the latter. Amongst the woody plant pollen that of shrubs is commonest in the arid belts but tree Pollen asserts itself as rainfall increases. The long-distance transport of pollen from one region to another does not affect the overall picture of the pollen-rain to any significant degree except in the areas bordering the Himalayas where pine pollen is over-represented. Cerealia type pollen usually occurs in areas where there is cereal cultivation but remains unrepresented in the very arid belt which is free of cereal agriculture. The pollen frequencies of Acacia, Prosopis, Zizyphus, Salvadora, Maytenus and the species belonging to Leguminosae, Labiatae and Geraniaceae are Under-represented. On the other hand 5% pine pollen occurs even as far as 600 km from the source.

Abstract: Pollen from fertile plants of corn (L.), with normal and restored male-sterile cytoplasm (), from various sources was assayed for germination on media containing pathotoxins extracted from infected leaves. The “pathotoxin” from race O inhibited neither germination nor growth of pollen from any of the or normal cytoplasm sources tested, as compared with controls. The pathotoxin from race T, however, inhibited germination and growth of pollen from and plants. Growth of pollen from , and normal cytoplasm sources was not inhibited by race T pathotoxin. The advantages of a pollen bioassay are discussed, as are possible special applications of this procedure.

Abstract: Mount Shisha Pangma,one of the high peaks of the central Himalayas,is about 8,012 m in height above the sea level.At the foot of its northern slope,there is a series of grayish yellow sandstones.Its exposed thickness is about 1,000 m. In the lower part of this sandstone series,a plant-bearing bed was discovered,in which were preserved leaves of evergreen oaks,Quercus semicarpifolia Sm.,Qu.cf. pannosa Hand.-Mzt.and Qu.cf.senecens Hand.-Mzt.,and a leaf of Cyperaceae . Under palynological investigation,a large amount of pollen grains of Cedrus(mostly of C.deodara Loud.)and Quercus has been observed,counted about 59.1% and 72.2% of the sporo-pollen contents of the upper and lower parts respectively.Besides,pollen of Abies,Picea,Pinus,Tsuga,Betula,Ericaceae,Labiatae and Cyperaceae and spores of Pteris,Polypodium and Selaginella were also observed.The sporo-pollen assemblages of the upper and lower parts of this series are quite similar.Owing to the increment of the amount of Abies pollen and the appearance of pollen grains of Picea,Pinus and some shrubby and herbaceous plants and spore of Selaginella,the flora of the later stage is more complex than that of the earlier,and the floral development of the later stage is more close to that in temperate climate rather than in subtropical.Anyway, the vegetation of that time is,as a whole,mainly represented by evergreen oak and Cedrus forests.The climate then seems to be rather mild and rainy. The flora is more close to the late Pliocene flora of Northwestern Yunnan,the middle-late Pliocene flora of south-eastern Europe and the early Pleistocene flora of Kashmir.The floral development of the said flora is quite identical with that of the Pliocene flora of Ebene(Sofia)of Bulgaria and that of the middle Pliocene flora of Cantal,southern France.The geological age of the present sandstone series is not earlier than the middle-late Pliocene. The climate of the fossil locality at present is very severe.The altitude is 5,700— 5,900 m,above the sea level.Plants can no longer live even in mid-summer.Both the macro-and micro-plant fossils indicate that the climate then was rather warm and the altitude was only 2,500 m above the sea level.Evidently,Mt.Shisha Pangma has up- lifted about 3,000 m.since late Pliocene.

Abstract: Application of gibberellic acid (GA3) to young plants of a single gene recessive stamenless-2 (sl2/sl2) mutant of tomato produced 'phenocopies' of the normal plants. Unlike the untreated sl2/sl2 mutant, flowers of GA3-treated plants bore no external ovules, possessed more yellow-pubescent stamens and fewer carpelloid stamens per flower, and produced laterally fused stamens. Stamen length at maturity was similar to normal flowers. In addition, viable pollen resembling the normal was produced in GA3-treated sl2/sl2 flowers. It was also found that GA3 was more effective than GA4+7 in inducing stamen development. Normal plants treated with GA3 produced multicarpellary and multilocular ovaries. Indoleacetic acid (IAA) induced the carpellization of stamens in sl2/sl2 flowers but had no apparent effect on the flowers of normal plants. It is proposed that added gibberellins promote maleness in systems where there is an inhibition or abnormality of stamen development, whereas they stimulate femaleness (possibly th...

Abstract: Summary The paper describes how absolute values of pollen deposition may be calculated from the results of volumetric sampling, supplemented by routine meteorological data. Daily mean values of concentration of birch pollen in the air at Lerwick, Shetland in 1970 and 1971 are presented, and these are used to calculate daily and annual depositions. The results are consistent with what is known about the occurrence of tree pollen in fossil and recent deposits in Shetland.

Abstract: INCREASING evidence indicates that late glacial and postglacial changes in climate and vegetation were essentially synchronous in the northern and southern hemispheres1,2. Pollen analysis of Quaternary deposits on subantarctic ocean islands should provide critical data, but has been inconclusive. The floras of most of the islands are depauperate and yield few identifiable sporomorphs. Species represented in the pollen rain are often of narrow geographical distribution and broad ecological tolerance, thus interpretation of changes in pollen and spore ratios and cross-correlations between islands are difficult3. These problems can be minimized by sampling in a location with a comparatively large flora, with many species widely distributed in the subantarctic, and with a variety of vegetation types. Kerguelen Islands fulfil these requirements.

Abstract: The formation of a tetrad of haploid microspores at the end of meiosis in the diploid microspore mother cells of seed plants marks the beginning of the short-lived gametophytic phase. The microspores mature into pollen grains which serve to transmit the sperms, via the pollen tube, to the female gamete. Biology of pollen occupies an important phase in the life cycle of seed plants, and its understanding is important not only for elucidating the basic biological problems of growth, differentiation, cell division, reproduction, etc., but also in the applied aspects of plant genetics, plant breeding, agriculture, and human allergies. The development, differentiation, and physiology of pollen grains are all complex processes, but much insight has been gained into the biology of pollen in the last two decades mainly by the use of advanced histochemical, immunological, cytoehemical, ultrastructural, autoradiographic, biochemical and tissue culture techniques. This is the New Biology of Pollen.