HEAT-PRODUCTION AND CROSS-POLLINATION OF THE ASIAN SKUNK CABBAGE SYMPLOCARPUS RENIFOLIUS (ARACEAE)

The Asian skunk cabbage Symplocarpus renifolius has an exothermic spadix on which about 100 flowers bloom in very early spring when effective pollinators such as bees and drone flies are inactive. This species is protogynous; female phase and male phase took 6.8 ± SD 5.8 days and 16.7 ± 5.7 days, respectively, with a short transitional phase of bisexuality (2.1 ± 0.9 days). The spadices produced heat 24 hours/day throughout female and bisexual phases, but temperature dropped quickly after the beginning of male phase. Although self-compatibility was expected from the flower structure, the basipetal flowering, and the absence of effective pollinators, bagging tests demonstrated that they rarely produce seeds without crossing. The spadices were visited by small numbers of invertebrates throughout the flowering season. Of these invertebrates, house flies, rove beetles, and mosquitos were the likeliest pollinators, since they are probably attracted both to the pollen produced in male phase and to the stench or carbon dioxide in female phase. On two female spadices with immature male flowers, we fortuitously collected a rove beetle and a mosquito that carried some pollen grains; these had to have been transported from other S. renifolius spadices. This infrequent and ineffective pollination appears to explain why as low as 13% of spadices set seeds in a natural population. We examine alternative hypotheses to explain production of heat in spadices of skunk cabbage.     

Relationships between the floral neighborhood and individual pollen limitation in two self-incompatible herbs

Local flower density can affect pollen limitation and plant reproductive success through changes in pollinator visitation and availability of compatible pollen. Many studies have investigated the relationship between conspecific density and pollen limitation among populations, but less is known about within-population relationships and the effect of heterospecific flower density. In addition, few studies have explicitly assessed how the spatial scales at which flowers are monitored affect relationships. We investigated the effect of floral neighborhood on pollen limitation at four spatial scales in the self-incompatible herbs Armeria maritima spp. maritima and Ranunculus acris spp. acris. Moreover, we measured pollen deposition in Armeria and pollinator visits to Ranunculus. There was substantial variation in pollen limitation among Armeria individuals, and 25% of this variation was explained by the density of compatible and heterospecific flowers within a 3 m circle. Deposition of compatible pollen was affected by the density of compatible and incompatible inflorescences within a 0.5 m circle, and deposition of heterospecific pollen was affected by the density of heterospecific flowers within a 2 m circle. In Ranunculus, the number of pollinator visits was affected by both conspecific and heterospecific flower densities. This did not, however, result in effects of the floral neighborhood on pollen limitation, probably due to an absence of pollen limitation at the population level. Our study shows that considerable variation in pollen limitation may occur among individuals of a population, and that this variation is partly explained by floral neighborhood density. Such individual-based measures provide an important link between pollen limitation theory, which predicts ecological and evolutionary causes and consequences for individual plants, and studies of the effects of landscape fragmentation on plant species persistence. Our study also highlights the importance of considering multiple spatial scales to understand the spatial extent of pollination processes within a population.     

DIRECT AND INDIRECT ESTIMATES OF NEIGHBORHOOD AND EFFECTIVE POPULATION SIZE IN A TROPICAL PALM,ASTROCARYUM MEXICANUM

To estimate the relative importance of genetic drift, the effective population size ???(N_e) can be used. Here we present estimates of the effective population size and related measures in Astrocaryum mexicanum, a tropical palm from Los Tuxtlas rain forest, Veracruz, Mexico. Seed and pollen dispersal were measured. Seeds are primarily dispersed by gravity and secondarily dispersed by small mammals. Mean primary and secondary dispersal distances for seeds were found to be small (0.78 m and 2.35 m, respectively). A. mexicanum is beetle pollinated and pollen movements were measured by different methods: a) using fluorescent dyes, b) as the minimum distance between active female and male inflorescences, and c) using rare allozyme alleles as genetic markers. All three estimates of pollen dispersal were similar, with a mean of approximately 20 m. Using the seed and pollen dispersal data, the genetic neighborhood area (A) was estimated to be 2,551 m². To obtain the effective population size, three different overlapping generation methods were used to estimate an effective density with demographic data from six permanent plots. The effective density ranged from 0.040 to 0.351 individuals per m². The product of effective density and neighborhood area yields a direct estimate of the neighborhood effective population size (N_b). N_b ranged from 102 to 895 individuals. Indirect estimates of population size and migration rate (Nm) were obtained using F_st for five different allozymic loci for both adults and seeds. We obtained a range of Nm from 1.2 to 19.7 in adults and a range of Nm from 4.0 to 82.6 for seeds. We discuss possible causes of the smaller indirect estimates of Nm relative to the direct and compare our estimates with values from other plant populations. Gene dispersal distances, neighborhood size, and effective population size in A. mexicanum are relatively high, suggesting that natural selection, rather than genetic drift, may play a dominant role in patterning the genetic variation in this tropical palm.     

Spatiotemporal mating pattern variation in a wind‐pollinated Mediterranean shrub

Spatiotemporal variation in mating patterns is poorly known in wind‐pollinated plant species. Here, we analysed mating patterns of the wind‐pollinated dioecious shrub Pistacia lentiscus by genotyping 904 seeds from 30 mother plants with eight microsatellite markers in a high‐density population in two consecutive flowering seasons. We found significant differences in some mating system estimates between years, particularly in the levels of correlated paternity. Overall, within‐mothers correlated paternity was higher in 2007 than in 2006 (r_pWM = 0.085 and 0.030), which translated into an effective number of fathers (N_ep) of 11.8 and 33.6 respectively. Using a smoothing interpolation technique, we show that the effective pollen cloud was spatially structured in patches of high‐ and low‐genetic diversity, which do not remain constant from year to year. In 2006, the among‐mothers correlated paternity (r_pAM) showed no trend with distance, suggesting no restriction of pollen dispersal. However, in 2007, r_pAM was greater than zero at short distances, revealing the existence of small‐scale patterns of pollen dispersal. The fact that the studied seasons were climatically homogeneous during the flowering time suggested that the observed differences might be ascribed to between‐year phenological variation of individuals in the studied population or other (unknown) factors. Numerical simulations, based on the real data set, indicated that the clumping of males and decreasing plant density, which is related to different types of pollen limitation, greatly increase correlated mating in this wind‐pollinated species, which is of relevance under the frame of the continuous anthropogenic habitat disturbance suffered by Mediterranean ecosystems.     

Ecological determinants of mating system within and between three Fagus sylvatica populations along an elevational gradient

Studies addressing the variation of mating system between plant populations rarely account for the variability of these parameters between individuals within populations, although this variability is often non‐negligible. Here, we propose a new direct method based on paternity analyses (Mixed Effect Mating Model) to estimate individual migration (m_i) and selfing rates (s_i) together with the pollen dispersal kernel. Using this method and the KINDIST approach, we investigated the variation of mating system parameters within and between three populations of Fagus sylvatica along an elevational gradient. Among the mother trees, s_i varied from 0% to 48%, m_i varied from 12% to 86% and the effective number of pollen donors (Nep_i) varied from 2 to 364. The mating patterns differed along the gradient, the top population showing higher m and lower s, and a trend to higher Nep than the bottom populations. The phenological lag shaped long‐distance pollen flow both within population (by increasing m_i at mother‐tree level) and between populations (by increasing m at high elevation). Rather than the mate density, the canopy density was detected as a major mating system determinant within population; it acted as a barrier to pollen flow, decreasing the proportion of long‐distance pollen flow and increasing s_i. Overall, the effects of ecological factors on mating system were not the same within vs. between populations across the gradient, and these factors also differed from those traditionally found to shape variation at range‐wide scale, highlighting the interest of multiscale approaches.     

Phylogeny and classification of Colocasiomyia (Diptera, Drosophilidae), and its evolution of pollination mutualism with aroid plants

Abstract Colocasiomyia, a moderate‐sized genus in the subfamily Drosophilinae, comprises seventy (twenty‐six described and forty‐four undescribed) species. Several Colocasiomyia species have evolved intimate mutualisms with specific host plants, especially of the family Araceae: the flies depend throughout the entire life cycle, oviposition, larval growth, pupation, and adult feeding and mating, on inflorescences of their host plants, and in turn act as species‐specific pollinators for their host plants. To understand the evolution of this mutualism between Colocasiomyia flies and their host plants, the phylogenetic relationships of this genus and some possibly related taxa are inferred from a cladistic analysis based on sixty‐two characters of adult morphology. We conclude that Colocasiomyia is polyphyletic, with the C. arenga species group clearly separate. Colocasiomyia without the arenga group (Colocasiomyia proper) is sister to all other studied drosophilines, whereas the arenga group is relatively derived within the Drosophilinae. Within Colocasiomyia proper, four clades are recognized, three of which correspond to previously proposed species groups: the cristata, toshiokai and baechlii groups. The other clade, C. sp.1 aff. nepalensis+C. sp.2 aff. nepalensis, is defined as a new species group. Relationships amongst the four clades and three independent species (C. micheliae, C. gigantea and C. sp.K1) remain almost unresolved, except for a sister group relationship between the toshiokai and baechlii groups. The classification of species groups in Colocasiomyia is revised by erecting two new species groups (crassipes and zeylanica groups) in addition to the three known (baechlii, cristata and toshiokai) groups. Revision of the arenga group, which should be removed from Colocasiomyia, is left for future studies. The evolution of host plant selection in Colocasiomyia is discussed by mapping host plant taxa (families, subfamilies and tribes) on the phylogenetic tree deduced from the cladistic analysis. Cohabitation in the same host inflorescence by a pair of species with microallopatric niche separation on the spadix is hypothesized to have evolved independently at least more than twice in Colocasiomyia.     

CONSTANCY OF POPULATION PARAMETERS FOR LIFE-HISTORY AND FLORAL TRAITS IN RAPHANUS SATIVUS L. II. EFFECTS OF PLANTING DENSITY ON PHENOTYPE AND HERITABILITY ESTIMATES

To determine the effect of growing conditions on population parameters in wild radish, (Raphanus sativus L.: Brassicaceae), we replicated maternal and paternal half-sib families of seed across three planting densities in an experimental garden. A nested breeding design performed in the greenhouse produced 1,800 F₁ seeds sown in the garden. We recorded survivorship, measured phenotypic correlations among and estimated narrow-sense and broad-sense heritabilities (h²) of: days to germination, days to flowering, petal area, ovule number/flower, pollen production/flower, and modal pollen grain volume. Survivorship declined with increasing density, but the relative abundances of surviving families did not differ significantly among densities. Seeds in high-density plots germinated significantly faster than seeds sown in medium- or low-density plots, but they flowered significantly later. Plants in high-density plots had fewer ovules per flower than those in the other treatments. Petal area and pollen characters did not differ significantly among densities. Densities differed with respect to the number and sign of significant phenotypic correlations. Analyses of variance were conducted to detect additive genetic variance (V_a) of each trait in each density. At low density, there were significant paternal effects on flowering time and modal pollen grain volume; in medium-density plots, germination time, flowering time and ovule number exhibited significant paternal effects; in high-density plots, only pollen grain volume differed among paternal sibships. The ability to detect maternal effects on progeny phenotype also depended on density. Narrow-sense h² estimates differed markedly among density treatments for germination time, flowering time, ovule number and pollen grain volume. Maternal, paternal and error variance components were estimated for each trait and density to examine the sources of variation in narrow-sense h² across densities. Variance components did not change consistently across densities; each trait behaved differently. To provide qualitative estimates of genetic correlations between characters, correlation coefficients were estimated using paternal family means; these correlations also differed among densities. These results demonstrate that: a) planting density influences the magnitude of maternal and paternal effects on progeny phenotype, and of h² estimates, b) traits differ with respect to the density in which heritability is greatest, c) density affects the variance components that comprise heritability, but each trait behaves differently, and d) the response to selection on any target trait should result in different correlated responses of other traits, depending on density.     

Selfing rates of pearl millet (Pennisetum typhoides Stapf and Hubb.) under natural conditions

The selfing rate of pearl millet (Pennisetum typhoides Stapf and Hubb.) has been determined under natural conditions. This species is said to be allogamous. Nine test plants homozygous for a particular allele on the alcohol dehydrogenase: ADH A locus (A₁A₁) were sown 2.5 m one from each other interspersed among 300 plants homozygous for the same locus (A₂A₂); these nine plants served as indicators of selfing. In the 20 spikes produced by these a test plants, the selfing rates varied between 2.2% and 21.7%. Selfmg rates were not significantly different within spikes of the same plant, except for one individual. There was no significant correlation between the rate of selfing and the density of the pollen shadow (estimated from the number of spikes producing pollen during the female phase of test plants) or variation in protogyny.     

Adaptations for insect‐trapping in brood‐site pollinated Colocasia (Araceae)

The Araceae include both taxa with rewarding and deceptive trap pollination systems. Here we report on a genus in which rewarding and imprisonment of the pollinators co‐occur. We studied the pollination of four species of Colocasia in Southwest China and investigated the morpho‐anatomical adaptations of the spathe related to the attraction and capture of pollinators. All four species were pollinated by drosophilid flies of the genus Colocasiomyia. The flies are temporally arrested within the inflorescence and departure is only possible after pollen release. Trapping of the flies is accomplished by the closure of the spathe during anthesis. Moreover, in two species the spathe is covered with papillate epidermal cells known to form slippery surfaces in deceptive traps of Araceae. However, in Colocasia the papillae proved not slippery for the flies. The morpho‐anatomical properties of the spathe epidermis indicate that it is an elaborate osmophore and serves for the emission of odours only. Despite its similarity to deceptive traps of other aroids, Colocasia and Colocasiomyia have a close symbiotic relationship, as the attracted flies use the inflorescence as a site for mating and breeding. The trap mechanism has presumably evolved independently in Colocasia and is supposed to facilitate more efficient pollen export.     

Pollen dispersal and genetic variation in an early‐successional forest herb in a peri‐urban forest

In order to conserve forest plant species under the particularly high constraints that represent urban surroundings, it is necessary to identify the key factors for population persistence. This study examined within‐ and between‐population pollen dispersal using fluorescent dye as pollen analogue, and genetic variation and structure using 15 allozyme loci in Centaurium erythraea, an insect‐pollinated, early‐successional forest biennial herb occurring in a peri‐urban forest (Brussels urban zone, Belgium). Dye dispersal showed an exponential decay distribution, with most dye transfers occurring at short distances (<15 m), and only a few long‐distance events (up to 743 m). Flowers of C. erythraea are mainly visited by Syrphids (Diptera) and small bees, which are usually considered as short‐distance pollen dispersers, and occasionally by bumblebees, which are usually longer‐distance pollen dispersers. Small and large dye source populations differed in dye deposition patterns. The populations showed low genetic diversity, high inbreeding coefficients (F_IS) and high genetic differentiation (F_ST), suggesting restricted gene flow, which can be expected for an early‐successional biennial species with a predominantly selfing breeding system and fluctuating population sizes. The positive relationship between recruitment rate and allelic richness and expected heterozygosity, and the absence of significant correlations between genetic variation and population size suggest seedling recruitment from the seed bank, contributing to maintain genetic diversity. Long‐distance dye dispersal events indicate pollinator movements along urban forest path and road verges. These landscape elements might therefore have a potential conservation value by contributing to connectivity of early‐successional species populations located in patchy open habitats.     

Isolation affects reproductive success in low-density but not high-density populations of two wind-pollinated <Emphasis Type="Italic">Thalictrum</Emphasis> species

Since pollen usually travels limited distances in wind-pollinated plant species, plants growing at low density may become pollen limited. We examined how local pollen availability and population density affect reproductive success in two wind-pollinated, dioecious species, Thalictrum fendleri and Thalictrum dioicum. Distance to the nearest flowering male, the number of flowering males within 2 m, and flower number on those males served as measures of local pollen availability. Increased distance from pollen donors reduced seed set in the lowest-density population of each species, but seed set in high-density populations was not correlated with local pollen availability. For plants in high- and low-density populations at similar distances from pollen donors, this distance only affected seed set in low-density populations. To ensure that differences in resource availability were not causing spurious correlations between seed set and plant density, we constructed low-density artificial arrays in populations of T. dioicum. In these, seed set decreased rapidly with increases in distance from pollen donors. Despite these effects, the density of males in a population was not correlated with average seed set in T. dioicum, and hand pollination in the T. dioicum populations also failed to increase seed set over natural levels. These results suggest that pollen receipt only limits seed set on isolated plants within low- density populations of T. dioicum and T. fendleri.     

ESTIMATION OF GENETIC NEIGHBORHOOD PARAMETERS FROM POLLEN AND SEED DISPERSAL IN THE MARINE ANGIOSPERM ZOSTERA MARINA L.

The relative importance of random genetic drift and local adaptation in causing population substructuring in plant species remains an important empirical question. Here I estimate the effective size of the genetic neighborhood, N_b, as a means of evaluating the likely role of genetic drift in creating genetic differentiation within a population of a marine plant, Zostera marina L. (eelgrass). Calculations of effective neighborhood size are based on field estimates of pollen and seed-dispersal distributions, an electrophoretic estimate of the mating system using open-pollinated progeny arrays, and determination of the effective density of reproductive individuals in the population. Neighborhood area calculated from the parent-offspring dispersal variances was equal to N_a = 524 m²; variance in the seed-dispersal distribution contributes more than twice as much as variance in pollen dispersal to N_a. Including an outcrossing rate slightly different from random, estimated neighborhood size for Z. marina is N_b = 6255. This estimate is one of the largest reported for plants or animals and indicates that genetic drift in small neighborhoods is highly unlikely to cause genetic substructuring in the study population. High gene-flow levels provided by the marine environment appear to prevent genetic isolation by distance among eelgrass patches, but the importance of drift through founder events in this population characterized by high patch turnover cannot be discounted and is the subject of ongoing study.     

Short-distance pollen dispersal and high self-pollination in a bat-pollinated neotropical tree

We investigated pollen dispersal and breeding structure in the tropical tree species Caryocar brasiliense Camb. (Caryocaraceae), using genetic data from ten microsatellite loci. All adult trees (101) within a patch of 8.3 ha were sampled, and from these adults 18 open-pollinated maternal progeny arrays were analyzed (280 seeds from 265 fruits). Most fruits presented only one seed (median equal to 1.000) and mean number of ripened seeds per fruit was 1.053 (SD = 0.828). Our results showed that C. brasiliense presents a mixed-mating system, with 11.4% of self-pollination, multilocus outcrossing rate of t _m = 0.891 ± 0.025, and high probability of full-sibship within progeny arrays (r _p = 0.135 ± 0.032). Outcrossing rate and self-pollination varied significantly among mother trees. We could detect a maximum pollen dispersal distance of ∼500 m and a mean pollen dispersal distance of ∼132 m. However, most pollination events (80%) occurred at distances less than 200 m. Our results also indicated that pollen dispersal is restricted to a neighborhood of 5.4 ha with rare events of immigration (∼1% N _e m = 0.35). C. brasiliense also presents a significant but weak spatial genetic structure (Sp = 0.0116), and extension of pollen dispersal distance was greater than seed dispersal (N _b = 86.20 m). These results are most likely due to the foraging behavior of pollinators that may have limited flight range. The highly within-population synchronous flowering, high population density, and clumped distribution reinforce pollinator behavior and affect residence time leading to a short-distance pollen dispersal.     

THE GENETIC CONSEQUENCES OF WORKER ANT POLLINATION IN A SELF-COMPATIBLE,CLONAL ORCHID

The self-compatible orchid Microtis parviflora is pollinated by the flightless worker caste of the ant Iridomyrmex gracilis. The orchid is clonal and forms small patches, usually less than 1 m², of disconnected individual ramets. Ant pollinators visited and revisited a limited proportion of available inflorescences, and 40% of all flower visits occurred within plants promoting self-pollination. Pollen labels indicated that self-pollination accounted for 51% of the pollen transfers, although pollen carryover extended beyond 16 flowers on 2 or 3 inflorescences. The distribution of ant movements between plants was leptokurtic with a mean of 12.4 ± 14.9 cm and a maximum of 89 cm, but a high proportion of movements were within clones accentuating the level of self-pollination. However, some pollen transfers between inflorescences of unlike genotypes contributed to a low incidence (max = 8%) of outcrossing. In 12 patches examined by electrophoresis, the density varied from 11 to 61 inflorescences per m² and a maximum of only 4 genotypes were detected. Electrophoretic analysis revealed populations were highly inbred: only 23% (N = 17) of the loci were polymorphic and the mean gene diversity h, was 2.7%. Heterozygotes were observed in only one population given a mean fixation index F, of 0.982. These results reflect the combined effects of restricted ant foraging and clonality. Nevertheless, while ant foraging was restricted, some outcrossing occurred and in the absence of clonality it is likely that ant foraging would have yielded a mixed mating system similar to those reported for a wide array of insect pollinators. Given the ability of ants to generate pollen flow, the reasons for the rarity of ant pollination appear to lie elsewhere.     

Genetic diversity of Lavandula multifida L. (Lamiaceae) in Tunisia: implication for conservation

The genetic variation within and among eight Tunisian natural populations of Lavandula multifida L., from different bioclimatic zones was assessed using random amplified polymorphic DNA (RAPDs). Of a total of 97 generated bands from seven selected primers, 84 bands were polymorphic. The genetic diversity within a population was high and varied according to the populations (0.308 < H’ < 0.459) without relationships to altitudes or pluviothermic indices of sites. The genetic differentiation among populations was high (G_ST = 0.395 and Φ_ST = 0.318). All population pairs were significantly differentiated. Among populations, within ecological groups genetic structure was high (0.219); whilst among them it was low (Φ_CT = 0.049; P < 0.05). The correlation between Φ_ST and geographic distance matrices among pairs of populations was not significant, suggesting that genetic connectivity between populations has a stochastic component at all spatial scales. The neighbour‐joining cluster analysis showed that individuals from each population clustered together. UPGMA cluster analysis showed that population groupings are not strictly in accordance with bioclimates or geographic location. The genetic differentiation in L. multifida could have occurred at local scales because of genetic drift. Efforts should be made to protect all populations. The maintenance of substantial population size should be initiated via fencing and controlling collection to restore the regeneration of populations.     

Genetic variation and population structure in Korean populations of eurya japonica (Theaceae)

We investigated levels of genetic diversity, population genetic structure, and gene flow in Eurya japonica, a widespread and broad-leaved evergreen dioecious tree native to Japan, China, Taiwan, and the southern and southwestern coast of the Korean Peninsula. Starch-gel electrophoresis was conducted on leaves collected from 1,000 plants in 20 Korean populations. All 12 loci examined were polymorphic in at least one population, and the mean number of alleles per locus was 3.79. In addition, mean observed population heterozygosity (Ho_p = 0.425), expected heterozygosity (He_p = 0.462), and total genetic diversity (H_T = 0.496) were substantially higher than average values for species with similar life history traits. Although significant differences in allele frequency were detected between populations at all loci (P < 0.001), <7% of the genetic variation was found among populations (F_ST = 0.069). There was a significant negative correlation between genetic identity and distance between populations (r = -0.341; P < 0.05), but this explained only a small amount of the diversity among populations. Indirect estimates of the number of migrants per generation (Nm) (3.37, calculated from F_ST; 3.74, calculated from the mean frequency of eight private alleles) indicate that gene flow is extensive among Korean populations of E. japonica. Factors contributing to the high levels of genetic diversity found within populations of E. japonica include large and contiguous populations, obligating outcrossing (dioecious plant), high fecundity, and long generation time. Occasional seed dispersal by humans and pollen movement by domesticated honey bees may further enhance gene flow within the species.     

Peptide growth factor phytosulfokine-α contributes to the pollen population effect

Density-dependent pollen germination and tube growth in vitro is a well-documented phenomenon, termed the pollen population effect, but far less is known about its molecular basis. We present evidence to support phytosulfokine-α [Y(SO₃H)IY(SO₃H)TQ; PSK-α] as a native bioactive factor contributing to this effect. Mature pollen grains of Nicotiana tabacum L. var.macrophylla were incubated in liquid medium for 2 h. Pollen germination frequency increased in a density-dependent manner from 625 to 46,000 grains/ml. Conditioned medium, obtained from the medium of pollen cultured at a density of 10,000 pollen grains/ml for 12 h, promoted the germination of pollen cultured at a low density (625 grains/ml). A rabbit antiserum against PSK-α specifically inhibited the promotive effect of conditioned medium. Quantification by enzyme-linked immunosorbent assay showed that the conditioned medium contained 0.4 nM of PSK-α. Exogenous PSK-α also stimulated pollen germination in the low-density culture. These results indicate that PSK-α is an important regulator involved in the pollen population effect. Received: 15 March 2000 / Accepted: 24 May 2000     

Hydrophilous pollination and reproductive morphology in the seagrassPhyllospadix scouleri (Zosteraceae)

In the seagrassPhyllospadix scouleri (Zosteraceae) floral aggregates (spadices) occur on short lateral axes produced by subordinate vegetative shoots. The filamentous pollen is both dispersed on the surface of the sea as well as below the surface. In surface pollination, snowflake-like search vehicles (pollen rafts) float and collide with any rigid female inflorescences that emerge from the water surface. In submarine pollination, collinear bundles of pollen grains are dispersed. Analysis of seed set indicates ovule position within the inflorescence to affect likelihood of fertilization in submarine, but not surface-pollinated inflorescences. Agamospermy appears to be unlikely, but the strongly female-biased shoot sex ratio remains to be explained.     

Novel polymorphic microsatellite loci and patterns of pollen-mediated gene flow in an ex situ population of <Emphasis Type="Italic">Eurycorymbus cavaleriei</Emphasis> (Sapindaceae) as revealed by categorical paternity analysis

Understanding the patterns of contemporary, pollen-mediated gene flow is of great importance for designing appropriate conservation strategies. In this study, ten novel polymorphic microsatellite loci were isolated for the rare dioecious tree, Eurycorymbus cavaleriei, and the patterns of pollen dispersal were investigated in an ex situ conserved population. A combination of microsatellite markers with high-collective exclusion power (0.932) was used to assign paternity to 240 seeds collected from eight maternal trees. The average effective pollen dispersal distance (δ) was 292.6 m and the frequency distribution of pollen movement suggested extensive pollen movement in the population. The effective pollen donors per maternal tree (N _ep) ranged from 5 to 10, and the most isolated maternal trees were observed with the largest number of N _ep = 10. Although a trend of near-neighbor mating was found in seven of eight maternal trees, no significant correlations were detected between the average effective pollen dispersal distance (δ) and the geographic distances (d1 and d2) between maternal and male trees. The increased average effective distance of pollen dispersal and number of N _ep for isolated maternal trees might be a compound consequence of low density and long-distance flight of pollinators of this species. The conservation implications of these results are discussed.