DEMOGRAPHIC GENETICS OF A PIONEER TROPICAL TREE SPECIES: PATCH DYNAMICS,SEED DISPERSAL,AND SEED BANKS

We consider whether changes in population-genetic structure through the life cycle of Cecropia obtusifolia, a tropical pioneer tree, reflect its gap-dependent demography and the role of evolutionary processes that are important for this species. We asked whether the spatial scale at which population-genetic subdivision occurs corresponds to the scale of habitat patchiness created by gap dynamics; whether patterns of seed dispersal and storage in the soil affect spatial genetic patterns; and whether spatial genetic patterns change through the species life cycle. We estimated Wright's F-statistics for six successive life-history stages for individuals grouped into subpopulations according to occurrence in natural gaps, physical proximity, or occurrence within large quadrats. For each life stage, F_ST-statistics were significantly higher when individuals were grouped by gaps, although concordant patterns across life stages for the three grouping methods were obtained. This supports the hypothesis that patchy recruitment in gaps or among-gap heterogeneity influences the species' genetic structure. F-statistics of seeds collected from females before dispersal (tree seeds), seed-rain seeds, soil seeds, seedlings, juveniles, and adults grouped by gaps, were, respectively: F_IT = 0.004, 0.160, 0.121, 0.091, –0.0002, –0.081; F_IS = –0.032, 0.124, 0.118, 0.029, –0.016, –0.083; and F_ST = 0.035, 0.041, 0.003, 0.063, 0.015, 0.002. Spatial genetic differentiation in rain seeds was not significantly lower than that of tree seeds. The loss of genetic structure in the soil seed bank, relative to that found in the seed rain may be due to sampling artifacts, but alternative explanations, such as microsite selection or temporal Wahlund effect are also discussed. If structure among soil seeds is unbiased, the peak in seedling F_ST may be due to microsite selection. F_IS of seeds in the rain and soil were significantly greater than zero. A Wahlund effect is the most likely cause of these positive F_IS values. Such fine-scale substructuring could be caused by correlated seed deposition by frugivores. The decrease in F_IS from seedlings to adults could result from loss of fine-scale genetic structure during stand thinning or from selection.     

GENETIC COMPARISONS OF SEED BANK AND SEEDLING POPULATIONS OF A PERENNIAL DESERT MUSTARD,LESQUERELLA FENDLERI

Soil seed banks may accumulate and store seed genotypes produced over many seasons. If germination and establishment of these soil seeds are influenced by seed genotypes, then seed bank and seedling populations may differ genetically. I compared the genetic structure of dormant but viable soil seeds of the desert mustard Lesquerella fendleri with the genetic structure of Lesquerella seedlings at the Sevilleta Long-Term Ecological Research Site. In 1991 and 1992, soil seeds and seedlings were mapped and genetically analyzed using starch gel electrophoresis. When data from all loci were lumped, there were highly significant differences in allele frequencies between soil seeds and seedlings at the population level (all plots) in both years, in all subpopulation (adjacent plots) comparisons in 1991, and three of five subpopulations in 1992. Differences at some individual loci were also detected in one or both years. Analysis of data pooled across both years revealed highly significant differences in the distribution of multilocus soil seed and seedling heterozygosity, but no significant differences in mean heterozygosity. F_st values showed small but statistically significant genetic differentiation within soil seeds and seedlings in both years. F_st values also showed significant genetic differentiation between these two groups at three of seven loci in 1991, and at one locus in 1992. Soil seeds and seedlings showed a general pattern of decreasing genetic relationship with distance, as estimated by the coefficient of coancestry analyses. In 1991, seedlings were roughly twice as genetically related to each other than were soil seeds at fine spatial scales (0–0.25 and 0.25–0.50 m). This study suggests that Lesquerella seedlings in this system represent a nonrandom genetic subset of the underlying Lesquerella seed bank. Such temporal genetic change may be an important yet frequently overlooked mechanism for generating population genetic structure.     

Optimal time of seedling emergence in a dune-population ofOenothera glazioviana

Oenothera glazioviana (= O. Erythrosepala) behaves as a facultative biennial in an infertile sand-dune system in Japan. In the field, seedlings emerge mainly in autumn following seed ripening in late summer with an additional small peak in spring. The effects of emergence time on subsequent survival and fecundity, and the ultimate numbers of seeds produced per emerging seedling (F _s) were examined by a demographic model. The model showed that seedlings emerging either in March or in September–October will attain higherF _s than those emerging in other months. If mortality in the seed bank is taken into account, autumn seems to be the best season for emergence to maximize the number of seedlings in the next generation.     

Fitness of resprouters versus seeders in relation to nutrient availability in two Plantago species

Two contrasting strategies of plants from disturbed areas are reported to depend on nutrient availability. Resprouters, investing into storage and capable of vegetative regeneration after disturbance, are predicted to be enhanced in nutrient poor environments. This contrasts to seeders, which invest preferentially into seed production and regenerating only from seeds, and are thought to prevail in nutrient rich environments. To test such predicted dichotomy, we set up an experiment with two facultative resprouters with contrasting nutrient demands and assessed the fitness of individuals regenerated from seeds and root fragments in differently productive environments. We hypothesized that 1) plants with higher nutrient demands have a higher fitness as seeders irrespectable of nutrient availability and/or 2) both species will have a higher fitness as resprouters under lower nutrient availability and as seeders when nutrient availability is higher. Nutrient availability was also manipulated prior to and after disturbance to evaluate the impact of changing nutrient availability on the strategy of resprouting. The results of our pot experiment with Plantago lanceolata and Plantago media supported the first but not the second hypothesis. Moreover, high nutrient availability prior to disturbance negatively affected resprouting success, but the growth and fitness of successfully regenerated individuals were enhanced under higher nutrient availability. We concluded that resprouting from roots after disturbance is affected by nutrient availability, but this effect considerably differs between individual life-history stages.     

Airborne‐pollen pool and mating pattern in a hybrid zone between Pinus pumila and P. parviflora var. pentaphylla

The reproductive isolation barriers and the mating patterns among Pinus pumila, P. parviflora var. pentaphylla and their hybrids were examined by flowering phenology and genetic assays of three life stages: airborne‐pollen grains, adults and seeds, in a hybrid zone on Mount Apoi, Hokkaido, Japan. Chloroplast DNA composition of the airborne‐pollen was determined by single‐pollen polymerase chain reaction. Mating patterns were analysed by estimating the molecular hybrid index of the seed parent, their seed embryos and pollen parents. The observation of flowering phenology showed that the flowering of P. pumila precedes that of P. parviflora var. pentaphylla by about 6 to 10 days within the same altitudinal ranges. Although this prezygotic isolation barrier is effective, the genetic assay of airborne‐pollen showed that the two pine species, particularly P. pumila, still have chances to form F₁ hybrid seeds. Both parental species showed a strong assortative mating pattern; F₁ seeds were found in only 1.4% of seeds from P. pumila mother trees and not at all in P. parviflora var. pentaphylla. The assortative mating was concluded as the combined result of flowering time differentiation and cross‐incompatibility. In contrast to the parental species, hybrids were fertilized evenly by the two parental species and themselves. The breakdown of prezygotic barriers (intermediate flowering phenology) and cross‐incompatibility may account for the unselective mating. It is suggested that introgression is ongoing on Mount Apoi through backcrossing between hybrids and parental species, despite strong isolation barriers between the parental species.     

Effect of recurrent selfing on inbreeding depression and mating system evolution in an autopolyploid plant

Genome duplication resulting in polyploidy can have significant consequences for the evolution of mating systems. Most theory predicts that self‐fertilization will be selectively favored in polyploids; however, many autopolyploids are outcrossing or mixed‐mating. Here, we examine the hypothesis that the evolution of selfing is restricted in autopolyploids because the genetic cost of selfing (i.e., inbreeding depression) increases monotonically with successive generations of inbreeding. Using the herbaceous, autotetraploid plant Chamerion angustifolium, we generated populations with different inbreeding coefficients (F= 0, 0.17 and 0.36) through three consecutive generations of selfing and compared their magnitudes of inbreeding depression in a common environment. Mating system estimates for four natural populations confirmed that tetraploid selfing rates (s_m= 0.25, SE = 0.02) are similar to those of diploids (s_m= 0.12, SE = 0.12; F_1,2= 1.34, P= 0.37) indicating that both cytotypes are predominantly outcrossing. Compared to an outbred control line, mean inbreeding depression for seed production, survival, and height (vegetative and total) in the inbred line differed among generations (inbreeding coefficients). Across all stages, inbreeding depression (relative to control) was positively related to generation (inbreeding coefficient). Although the initial costs of inbreeding in extant and newly synthesized polyploids may be low compared to diploids, the monotonic increase in inbreeding depression with repeated inbreeding may limit the extent to which selfing variants are favored.     

Why be a honeyless honey mesquite? Reproduction and mating system of nectarful and nectarless individuals

Populations of Prosopis glandulosa var. torreyana in the Chihuahuan desert have a fixed dimorphic system of nectar production in which half the individuals produce nectar (are nectarful) and the other half are nectarless. We analyzed the impact of nectar production on different estimates of fitness, comparing nectarful against nectarless individuals in size, mating system, seed traits, and fruit set in a 1-ha scrubland. Of the reproductive individuals (358), 46% were nectarful and 54% were nectarless. Neither tree size nor flowering phenology differed between nectar morphs. Fixation indices (F) for both progeny (F = −0.2) and adults (F = −0.45) were negative, and high heterozygosities were found in adults and progeny (H = 0.45). No differences were found between nectar morphs for F, H, and single (t_s = 1.1) and multilocus (t_m = 1.03) outcrossing rates. Controlled pollinations showed differences between selfing and control treatments with no differences between nectar morphs. Nectarless individuals produced significantly more pollen grains than did nectar producers, but all other measured floral traits showed no differences. Nectarful trees were visited by pollinators 21 times more often and had a significantly higher overall fruit set than did nectarless trees. No differences between nectar morphs in seed mass or in percentage seed germination were found, but heavier seeds tended to have higher heterozygosities. Both morphs had similar success as females, but nectarless trees had ∼7% higher male function. We discuss three possible scenarios for the evolution of the fixed dimorphism in nectar production, two involving unstable phases (substitution of one morph by the other, and evolution towards dioecy) and one stable scenario (maintenance of the dimorphic system).     

Selfing and inbreeding depression in seeds and seedlings of Neobalanocarpus heimii (Dipterocarpaceae)

We evaluated the degree of selfing and inbreeding depression at the seed and seedling stages of a threatened tropical canopy tree, Neobalanocarpus heimii, using microsatellite markers. Selection resulted in an overall decrease in the level of surviving selfed progeny from seeds to established seedlings, indicating inbreeding depression during seedling establishment. Mean seed mass of selfed progeny was lower than that of outcrossed progeny. Since the smaller seeds suffered a fitness disadvantage at germination in N. heimii, the reduced seed mass of selfed progeny would be one of the determinants of the observed inbreeding depression during seedling establishment. High selfing rates in some mother trees could be attributed to low local densities of reproductive individuals, thus maintenance of a sufficiently high density of mature N. heimii should facilitate regeneration and conservation of the species.     

Regenerative traits of tree species in a deciduous forest of northeastern North America

I studied the seed rain, seedling bank, newly emerged seedlings, and seed bank of tree species in a deciduous forest of northeastern North America (Québec, Canada). The main objective was to determine to what extent the tree species present on the site differed in some of their regenerative traits. All the species present on the quadrats as mature trees contributed to the annual seed rain, although seed rain was strongly dominated by yellow birch and sugar maple. The seedling bank was mostly composed of red maple and sugar maple, but inputs from the 1988 cohort were mostly red maple and yellow birch seedlings. Seedling mortality was high in general (from 60% to 100%), but sugar maple had the lowest mortality of the tree species present. Two species maintained a seed bank on the study site: yellow birch and american basswood. Some species not present on the quadrats, but having populations nearby, disseminated seeds onto the quadrats (i.e. white ash and grey birch). Other species, not present on the quadrats as mature individuals, had seeds in the seed bank although they did not participate in the seed rain, did not maintain a seedling bank, and did not establish seedlings in 1988 (i.e. pin- cherry and Rubus spp.). Red maple, sugar maple, and yellow birch showed contagion in the dispersion pattern of their seeds in the seed rain (and the seed bank for yellow birch), and of their seedlings in the seedling bank and the 1988 cohort. Seedlings of american beech and american basswood of the seedling bank and the 1988 cohort were randomly dispersed over the quadrats, although their seeds in the seed rain had a contagious dispersion pattern. It has often been proposed in the literature that regenerative traits come only in “compatible assemblages” and that, consequently, species may be grouped according to their strategy of regeneration: e.g. species that maintain a seed bank vs those with a seedling bank; species with either a seed or a seedling bank vs those with the ability to sprout. In the community studied, it appears that regenerative traits do not come in neatly packaged exclusive assemblages, although the species do seem to possess different combinations of regenerative traits. However, yearly variations in seed production, seed germination, and seedling survival may render difficult the identification of strategies of regeneration.     

Heterozygote excess through life history stages in Cestrum miradorense Francey (Solanaceae), an endemic shrub in a fragmented cloud forest habitat

Comparisons of genetic diversity and population genetic structure among different life history stages provide important information on the effect of the different forces and micro‐evolutionary processes that mould diversity and genetic structure after fragmentation. Here we assessed genetic diversity and population genetic structure using 32 allozymic loci in adults, seeds, seedlings and juveniles of eight populations of the micro‐endemic shrub Cestrum miradorense in a highly fragmented cloud forest in central–eastern Mexico. We expected that due to its long history or rarity, this species may have endured the negative effects of fragmentation and would show moderate to high levels of genetic diversity. High genetic diversity (H_e = 0.445 ± 0.03), heterozygote excess (F_IT = ?0.478 ± 0.034, F_IS = ?0.578 ± 0.023) and low population differentiation (F_ST = 0.064 ± 0.011) were found. Seeds had higher genetic diversity (H_e = 0.467 ± 0.05) than the later stages (overall mean for adults, seedlings and juveniles H_e = 0.438 ± 0.08). High gene flow was observed despite the fact that the fragmentation process began more than 100 years ago. We conclude that the high genetic diversity was the result of natural selection, which favours heterozygote excess in all stages, coupled with a combination of a reproductive system and seed/pollen dispersal mechanisms that favour gene flow.     

格氏栲天然林林窗和林下种子散布及幼苗更新研究

以格氏栲(Castanopsis kawakamii)天然林为研究对象,探讨了林窗和林下格氏栲种子雨、种子库的分布特征及幼苗更新状况。结果表明:林窗种子雨总量和完好种子密度高于林下,未成熟种子比例低于林下;林窗和林下种子雨高峰期掉落数量分别占种子雨总量的77.13%和74.5%;林窗种子库储量低于林下,种子库中以全食或捡拾种子比例最高,其中种子库储量垂直分布表现为枯落物层(约占2/3)>腐殖质层(0～5 cm)(约占1/3)>心土层(5～10 cm)(小于1%),以格氏栲种子占绝对优势;格氏栲从种子到幼苗的转化率低,林窗中格氏栲完好种子密度与幼苗密度均高于林下。林窗微生境提高了种子散布过程中格氏栲成熟种子的密度和比例,有利于促进格氏栲的幼苗更新,表明林窗在格氏栲种群恢复过程中扮演着重要角色。     

Effect of inbreeding on production of filled seed in Pinus radiata — experimental results and a model of gene action

Summary Pinus radiata D. Don was inbred to different degrees, commencing with a founder population of 8 clones. Yield of filled seed was determined for each mating type. Mean yields (%), relative to outcrossing, were: half-sib 102; full-sib 97; S₁ 43; S₂ 42. An epistatic model was developed to predict the empty seed yield following selfing, assuming that homozygosity for several recessive co-lethals is necessary to kill an embryo. Calculations were then extended to predict the yields following different degrees of inbreeding. The proposed model gave a better fit to present results, and to other published results of similar investigations, than did an alternative based on action of independent recessive lethals. Implications for breeding and seed production strategy are discussed.     

Pollen Dispersal Between Isolated Trees in the Brazilian Savannah: A Case Study of the Neotropical Tree Hymenaea stigonocarpa

This case study examines the pollen dispersal distance, pollen dispersal patterns and intra‐family genetic structure for isolated trees in pastures of the bat‐pollinated Neotropical tree species Hymenaea stigonocarpa using six microsatellite loci and parentage analysis. The sampling included 28 grouped trees (referred to as the population) and six isolated trees in pastureland along a highway in Mato Grosso do Sul State, Brazil. From the population, we sampled 137 seeds from 12 seed‐trees, and from the isolated trees, we sampled 34 seeds from two seed‐trees. The results showed that pollen was dispersed over long distances (reaching 7353 m) and therefore the spatially isolated trees were not reproductively isolated. The pollen immigration rate in the population was also high (31%). Isolated trees presented a higher selfing rate (s=26%) than trees in the population (s=12%), suggesting that the spatial isolation of the trees increased selfing. However, selfing was responsible for only 30 percent of the inbreeding in offspring and mating among relatives was 70 percent. In the population, excluding selfing, ca 72 percent of the pollen was dispersed over distances <1000 m (average: 860 m). For the two isolated seed‐trees, excluding selfing, the average pollen dispersal distance was 5229 m. The results demonstrate that although pollen can be dispersed over long distances for H. stigonocarpa isolated trees, a high percentage of pollen comes from the same tree (selfing) and mating was correlated. Consequently, seeds must be collected from a large number of seed‐trees for conservation purposes.     

Genetic variability in Plantago species in relation to their ecology

Summary Populations of the three Plantago species P. major, P. lanceolata and P. coronopus were scored for allozyme variability. They showed normal variability levels compared with other, similar plant species. Differentiation among populations appeared to be rather low in comparison with other species, probably due to a considerable amount of long distance seed transport. In order to be able to make an analysis of small-scale gene flow, all (sub)populations were critically checked for the existence of population structure equilibrium. The allozyme variation was tested for neutrality by testing homogeneity of F values among loci: between populations (Lewontin-Krakauer test) and within populations. No systematic deviations from the prediction of the neutral theory could be established. From the population structure analyses, little gene flow in the species P. major (with high selfing levels) and P. coronopus could be concluded, whereas P. lanceolata showed relatively high levels of gene flow. The degree of homozygosity in the latter species was too high to explain with the available data. In P. coronopus, on the other hand, an unusually high number of heterozygotes were observed.Grassland Species Research Group Publication No. 73     

The evolution of intrinsic reproductive isolation in the genus Cakile (Brassicaceae)

In theory, adaptive divergence can increase intrinsic post‐zygotic reproductive isolation (RI), either directly via selection on loci associated with RI, or indirectly via linkage of incompatibility loci with loci under selection. To test this hypothesis, we measured RI at five intrinsic post‐zygotic reproductive barriers between 18 taxa from the genera Cakile and Erucaria (Brassicaceae). Using a comparative framework, we tested whether the magnitude of RI was associated with genetic distance, geographic distance, ecological divergence and parental mating system. Early stages of post‐zygotic RI related to F₁ viability (i.e. initial seed set) tended to be stronger than later stages related to F₁ fecundity (i.e. flower number, fruit number). Mating system significantly influenced early stages of RI, such that RI was lowest when the mother was selfing and father was outcrossing, consistent with an imbalance between sink strength and resistance to provisioning. We found little evidence that adaptive divergence accelerates the evolution of intrinsic post‐zygotic RI, consistent with a nonecological model of evolution that predicts the nonlinear accumulation of RI and RI asymmetry with time (i.e. genetic distance), irrespective of adaptive divergence. Thus, although certain aspects of ecological divergence do not appear to have contributed strongly to the evolution of RI in this system, divergence in mating system actually reduced RI, suggesting that mating system evolution may play a significant role in speciation dynamics.     

Little to no inbreeding depression in a tapeworm with mixed mating

Meta‐studies on hermaphrodites have found a negative relationship between primary selfing rates and levels of inbreeding depression (ID) and, thus, generally support purging in inbred systems. However, in plants, high among‐taxa variance in ID results in no difference in the mean ID between outcrossing and mixed‐mating taxa. Selective interference likely explains high ID among mixed‐mating taxa, whereas low levels of ID among mixed‐mating taxa are not as stressed. Among animal hermaphrodites, primarily molluscs, there are little data on mixed‐mating systems. To fill a taxonomic and mating system gap, we tested for ID in a mixed‐mating tapeworm, Oochoristica javaensis. We provide a direct estimate of ID across infection of an intermediate host by comparing selfing rates at two life history stages. We found little to no evidence for ID, and the level of ID falls in line with what is reported for highly selfing species even though O. javaensis has mixed mating. We discuss this result within the context of kin mating in O. javaensis. Our results emphasize that primary selfing rates alone may be insufficient to classify the inbreeding history in all species when testing for a relationship to ID. Mixed‐mating taxa, and possibly some outcrossing taxa, may exhibit low levels of ID if biparental inbreeding is also driving purging. We advocate that ID studies report estimates of inbreeding history (e.g. F_IS or identity disequilibrium) from nature‐derived adult samples to provide context rather than relying on primary selfing rates alone.     

Quantitative genetic analysis of embryo heterosis in faba bean (Vicia faba L.)

Seeds, i.e. embryos, may be genetically different from either of their parents and moreover may express their own heterosis. The objective was to genetically analyse embryo heterosis for their own weight (i.e. seed weight) in comparison with their seedlings’ heterosis, taking the large-seeded crop (Vicia faba L.) as model. A specific diallel mating scheme was used, based on four parental lines, creating 76 seed genotypes in generations P, F₁, F₂ and BC. Mature seed weight was assessed for these embryo genotypes in 3 years at one German location, and young plant biomass yield of seedlings emerging from these seeds in two greenhouse experiments. The quantitative genetic analysis showed an average of 10.6% mid-parent heterosis for mature seed weight and 14.5% mid-parent heterosis for juvenile biomass. In both traits, the embryos contributed markedly and significantly via their own genes to the genetic variation. For mature embryo weight heterosis, apparently the parental difference in seed weight was decisive, whereas for juvenile biomass heterosis, genetic unrelatedness of parents had priority.     

Breaking of Seed Dormancy by Nitrate as a Gap Detection Mechanism

Germination of Planlago lanceolata seeds buried in a chalk grasslandwas higher in bare soil than in vegetated soil, and measurementof soil nitrate concentrations showed that they were high enoughto account for this stimulation. When seeds of P. lanceolatawere sown in pots of soil with or without plants, and wateredwith nutrient solution containing either no nitrate, or 14 mMnitrate (sufficiently high that not all nitrate was absorbedby the plants), the presence of plants inhibited germinationonly when the nutrient solution contained no nitrate. It wasconcluded that breaking of seed dormancy by nitrate can functionas a gap detection mechanism if nitrate concentrations in baresoil are high enough to break seed dormancy, but are too lowto break dormancy when vegetation is present Plantago lanceolata, seed, dormancy, germination, nitrate, gap detection     

How does population genetic diversity change over time? An experimental seed bank study of Atriplex tatarica (Chenopodiaceae)

Atriplex tatarica is an annual, early successional, facultative halophilic species of frequently disturbed human-made habitats in Central and Eastern Europe. We investigated to what extent the plants grown from seeds extracted from soil seed bank differed genetically to mature aboveground plants in experimental populations of A. tatarica over two successive years. At each of five plots 50 aboveground plants and 50 plants extracted from seeds stored in soil were assayed for allozyme analysis in 2003 and 2004. At the start of experiment, we introduced 1000 seeds of the study species into each of five experimental plots. While the species dominated in all of the experimental plots in the first year, the second year A. tatarica coverage decreased dramatically. Overall allele frequencies of soil seeds and mature plants showed significant differences between life history stages in both years, but not within years in soil seeds as well as mature plants stages. While mature plants showed a significantly greater amount of single and multilocus heterozygosity in both consecutive years, comparison between years did not yield any significant differences. In the same way, despite a relatively large seed bank the species population genetic parameters, i.e. allelic richness (A), observed heterozygosity (H_o), gene diversity (H_s), inbreeding coefficient (F_IS) and fixation index (F_ST), did not change over years between as well as within life history stages. The soil seeds and mature plants significantly differed in H_o, H_s and F_IS, while the A and F_ST were not significantly different between life history stages.     

Survivorship and growth of seedlings of four dipterocarp species in a tropical rainforest of Sarawak,East Malaysia

Survivorship and growth of seedlings of four dipterocarp species (Dipterocarpus actangulus, D. globosus, Dryobalanops aromatica, Dryobalanops lanceolata) were studied for 2.5 years in a mixed dipterocarp forest in Sarawak, East Malaysia. Predispersal seed predation rates were larger forD. globosus (75%) thanD. lanceolata (27–34%) andD. aromatica (18–26%). Less than 20% of the twoDryobalanops seeds were damaged by vertebrates after seed dispersal. During the period from seed dispersal to the time when the seedlings had shed cotyledons, more dispersed seeds died in the twoDipterocarpus (ca 90%) than the twoDryobalanops (ca 60–70%). The major mortality factors during this period were uprooted and seed/seedling predation by insects or vertebrates. After the seedlings shed cotyledons, all species showed constant mortality rates of 34, 15–16, 17 and 6%/year forD. actangulus, D. lanceolata, D. aromatica andD. globosus, respectively, in the forest understorey. Mortality was lower in less shaded conditions than in more shaded ones forD. aromatica andD. actangulus, but not significantly different forD. lanceolata andD. globosus. A majority of dead seedlings were killed by fallen branches or were found standing with wilted leaves, probably due to water stress. No significant correlation was found between seed/seedling mortality and distance from mother trees or the initial density of seeds/seedings for all species. The mean leaf production was positively correlated with the estimated diffuse light factor of their habitats for each species.