Effects of smoke,heat and charred wood on the germination of dormant soil‐stored seeds from a Eucalyptus baxteri heathy‐woodland in Victoria,SE Australia

The effects of dry heat, wet heat, charred wood and smoke on the germination of dormant soil‐stored seeds from a Eucalyptus woodland in western Victoria were tested by using a glasshouse seed‐bank germination experiment. Seedling density, species richness and species composition were compared between replicated treated and control samples. A total of 5922 seedlings, comprising 59 plant species, was recorded from the soil samples over a period of 150 days. While a few species dominated (including Centrolepis strigosa, Wahlenbergia gracilenta and Ixodia achillaeoides), 26 species were represented by fewer than five seedlings and 18 species were restricted to single treatment types. With the exception of charred wood, all treatments led to a significant increase in seed germination relative to the control. The highest number of germinants was obtained for the smoke treatment, with a mean (± SE) of 12 547 ± 449 seedlings m^–2. Heat treatments yielded intermediate densities, with means (± SE) varying between 7445 ± 234 and 9133 ± 445 seedlings m^–2. In comparison with the estimates of seed‐bank sizes from other fire‐prone ecosystems, these densities are high. Species richness differed significantly among treatments. Highest mean richness was recorded in the smoke treatment and lowest for the control and charred wood treatments. There were significant differences in seed‐bank species composition between treatment types based on analysis of similarity (Anosim) using Bray–Curtis similarity. While heat was a specific requirement for triggering germination in hard‐seeded species (e.g. Fabaceae), smoke was the most effective trigger for species from a broad range of other families. The potentially confounding effect of physical and chemical mechanisms of germination stimulation in heated bulk soil samples is raised as an issue requiring further investigation in relation to the role of smoke as a germination trigger.     

Smoke and heat effects on soil seed bank germination for the re‐establishment of a native forest community in New South Wales

The effects of plant‐derived smoke and of heat on the emergence of seedlings from seeds were assessed. Seeds had been stored in forest topsoil used for mine site rehabilitation. The study was carried out in a dry sclerophyll, spotted gum (Corymbia maculata), forest community at the Mount Owen open‐cut coal mine in the Hunter Valley region of New South Wales. Samples of the surface 2.5 cm of topsoil were either exposed to cool smoke from eucalypt foliage for 60 min, heated to 80°C, or left untreated. Seedling emergence from the seed bank in this soil was then monitored in a glasshouse. Within the first month, smoke alone promoted a 4.3‐fold increase in the density of seedlings relative to control. There were 540 emergents per m² in the control and 2309 per m² in the smoke treated topsoil. Many annual and perennial herbs emerged but grasses responded most strongly to smoke. Germination in seven of the 20 grass species was promoted by smoke. Smoke promoted the germination of some introduced species as well as native species, and accelerated the rate at which seedlings emerged, although these differences sometimes declined with time. Heat also stimulated germination but smoke and heat stimuli appeared to be complementary in their promotion of seedling emergence from the topsoil seed bank. Each treatment increased the density of different species, enhanced the species richness of different components of the seed bank, and had different effects on the rate of emergence. The results suggest that increased seed germination in the field immediately following a moderate intensity fire may sometimes be the result of smoke stimulation and sometimes the result of heat stimulation of the soil seed bank. These findings may have important implications for minesite revegetation programs where topsoils are replaced after mining and rapid germination of seeds stored in these soils is required during short periods when conditions are favourable for germination.     

Regeneration of monsoon rain forest in northern Australia: the dormant seed bank

Abstract. As part of a wider study examining regeneration pathways in monsoon rain forest vegetation in northern Australia, the dormant component of the soil seed bank was assessed by storing soil samples for over six dry season months, before watering in shade-house trials. Six soil samples were collected from each of 34 sites broadly representative of the range of regional monsoon rain forest vegetation. Four floristic seed bank groups were derived through TWINSPAN classification. Mean group densities of germinants ranged from 25–144/m². Dormant seed banks were least dense, and most sparsely distributed, in sandy soils. Seed bank samples were dominated by woody pioneer monsoon rain forest species, especially figs; exotic weeds and savanna taxa (e.g. Poaceae) were relatively more common at seasonally dry sites. Dormant seed banks comprised species mostly present in the standing vegetation, although a small number of germinants represented species not growing at half the sites. Regeneration of woody pioneers from dormant seed banks is least likely to be of importance on infertile, seasonally dry sites.     

Smoke Enhanced Seed Germination for Mine Rehabilitation in the Southwest of Western Australia

Cool smoke treatments were applied to unmined Eucalyptus marginata (jarrah) forest soils, rehabilitated bauxite mine soils, and broadcast seed to determine if enhancement in germination could be effected with a view toward maximizng the establishment of species in bauxite mines in Western Australia. Forest sites showed a 48-fold increase in total germinants from the soil seed bank when treated with aerosol smoke. Newly returned bauxite mine soils showed a greater than threefold increase in total germinants after the same treatment. There were also significant increases in the number of species germinating in response to the aerosol smoke treatment in both the forest and the mined soils. Similarly, application of smoked water to the soil seed bank in previously mined sites elicited a significant positive germination response, increasing total germinants and species numbers by 56 and 33%, respectively. Treatment of mixed seed lots with aerosol smoke before broadcast resulted in highly significant improvement in germination when compared to untreated seed. Both total number of germinants, and number of species emerging from mined sites were positively influenced (85% and 34% increases, respectively). Ten target species were used to determine the relative effectiveness of different methods of smoke treatment on the germination of broadcast seed. Nine of the species involved displayed a promotive effect with at least two of the treatments. Generally, however, aerosol smoking of seed before broadcast proved to be the more effective approach. As a result of these findings, all broadcast seed for use in Alcoa's bauxite mined areas in the southwest of Western Australia is now routinely smoke treated before application.     

Fire‐related cues and germination from the soil seed bank of senescent remnants of mallee vegetation on Eastern Kangaroo Island

Plant communities dominated by narrow‐leaved mallee (Eucalyptus cneorifolia) are almost entirely confined to north‐eastern Kangaroo Island, South Australia, an area which has been extensively cleared for agriculture. Consequently, surviving examples consist mostly of small remnants which are thought to be senescent due to the exclusion of fire. This senescence is associated with the loss of many native understory species. Prescribed burns have been suggested as a management tool to stimulate the restoration of native plants from the soil seed bank; however, no seed bank studies have previously been conducted on Kangaroo Island and the seed bank literature usually focuses on particular species rather than on plant communities. We conducted an experiment to investigate the effects of the fire‐related cues heat and smoke on the germination of plants from the seed bank in soil sampled from 10 long‐ungrazed narrow‐leaved mallee sites on Kangaroo Island. Eighty trays of soil were monitored in a controlled glasshouse for five months after being subjected to heat and/or smoke treatments. The overall number of native, but not exotic, plant species germinating from the soil seed bank was significantly increased by all three fire‐related treatments (heat, smoke and heat plus smoke) compared with the control (no fire‐related treatment). Different plant life forms exhibited varying responses to heat and smoke treatments. The results of this study illustrate that the application of fire‐related treatments to soil seed banks in controlled glasshouse conditions can stimulate the recruitment of native species, including several species of conservation concern. These findings also indicate the potential of using these treatments for the ex situ germination of fire dependent species for revegetation purposes and indicate aspects of prescribed burns that may be important for restoring different components of native vegetation.     

Composition and dynamics of wetland seed banks on Mount St. Helens, Washington, USA

The 1980 eruption of Mount St. Helens deposited new substrates upon which wetlands are now developing. We collected soil from five wetlands south of Spirit Lake, Mount St. Helens, Washington, to examine the seed banks. Seedling emergence density in the top 5 cm was highly variable, and ranged from (mean±s.d.) 15,700±15,200 to 38,000±31,500 per m². Seedling emergence from soil at 5 to 10 cm depth varied from 800±600 to 18,000±24,800 per m², and averaged 1/3 as many seeds as the surface. The high proportion of buried seeds may be due to continuing deposition of upland sediments. We identified 17 taxa from the wetland soil samples withEpilobium ciliatum, Juncus bufonius, Agrostis exarata andJuncus ensifolius being the dominant species. Wetland vegetation was dominated byEquisetum arvense, Salix sitchensis, Typha latifolia and the four dominant species in the seed bank. Correlations between vegetation and seed bank were not significant for four of the five wetlands, and the dominance byTypha latifolia in two wetlands contributed to the low correlations. Thus, vegetation of these early successional wetlands generally was not similar to seed bank composition. Canonical correspondence analysis was used to relate soil parameters to vegetation and seed bank data. The vegetation was correlated primarily with soil particle size and secondarily to water temperature, while the seed bank was correlated to pH, texture and temperature.     

Soil seed banks in different environments: initial forest,mature forest,Pinus and Eucalyptus abandoned stands

Abstract

Different environments (initial forest, mature forest, Pinus and Eucalyptus stands) found in Seasonal Semideciduous Forest fragments affect the density of viable seeds, as well as the floristic similarity, diversity, and richness of tree species in soil seed banks. This hypothesis was tested in the current study. Soil seed bank samples were collected in the aforementioned environments during rainy and dry seasons, and taken to a shade house, where they remained under favorable seed germination conditions. Tree seedling emergence was measured, and sample species were identified every 15?days, for six months, in each sampling period. In total, 97 individuals m^?2 and 23 species emerged in all environments and periods. The highest density of viable seeds of tree species in the soil seed bank was found in the initial forest stretch, mature forest stretch and abandoned Eucalyptus stand. Only the Pinus stand seed bank in the dry season had different floristic and lower viable seed density than the mature forest seed bank. Thus, all environments, except the abandoned Pinus stand, can preserved Seasonal Semideciduous Forest fragments.     

Spatio-temporal variation in a seed bank of a semi-arid region in northeastern Brazil

This study aimed to evaluate variations in the seed bank within a 3-year temporal series in order to answer the following questions: 1) Does the seed bank's species richness and seed density differ among climatic seasons and between years? 2) Are there differences in the richness and density of seed banks between the litter and mineral soil? 3) Can the seed bank's species richness and seed density be explained by characteristics such as the previous year's precipitation and soil depth (litter or mineral soil)? The samples were collected from litter and mineral soil (0–5 cm), in 210 sub-plots, during the dry and rainy seasons of each year (August 2005 through February 2008). Overall, 79 species were recorded. On average, 1 168, 304 and 302 seeds.m⁻² were recorded in the seed bank during years I, II and III, respectively. This study showed that the Caatinga's seed bank is rich in herbaceous species, yet species' density and richness are low in the litter. Furthermore, about 43% of the variation in species richness and density was explained by soil depth (litter and mineral soil) and previous years' rainfall.     

Heat shock increases the reliability of a temperate calcareous grassland seed bank study

Reliable assessment of the density and species richness of the viable seeds in the soil is essential to estimate the probability of successful restoration of a particular plant community. Since temperate calcareous grasslands contain several thermophilous species typical of fire prone habitats, heat shock can be expected to break dormancy of target species for calcareous grassland restoration. We tested the effect of heat shock on the estimated density and species richness of seeds in soil samples. Heat shock treatments between 70 and 110°C resulted in both a higher estimated seed density and estimated species richness, in particular of specialist calcareous grassland species. Moreover, 25 species germinated exclusively from the heat-treated samples. This indicates that several species would have been missed by handling soil seed bank samples following the standard germination protocol, resulting in an underestimation of the restoration potential.     

天山中部天山云杉林土壤种子库年际变化

通过定点采样,采用萌发法对天山中部天山云杉(Picea schrenkiana Fisch.et Mey.)近熟林(101-120a)和成熟林(121-160a) 2004-2011年(8a)土壤种子库物种组成、种子密度的年际变化和不同间隔年限土壤种子库物种组成的相似性进行了分析.结果表明:(1)土壤种子库中共萌发鉴定出种子植物87种,隶属29科70属,其中乔木种子植物2种,灌木种子植物2种,草本种子植物83种,土壤种子库中草本植物种子密度远远大于木本植物种子密度;8个采样年份土壤种子库恒有种仅有6种;(2)土壤种子库种子密度及其中天山云杉种子密度存在巨大的年际变动,且不具有同步性;土壤种子库种子密度最大(2009年)值为(953.75±66.12)粒/m2,最小(2008年)值为(186.50±20.37)粒/m2,其中天山云杉种子密度最高(2006年)达到(584.50±53.58)粒/m2,最低(2005年)仅有(0.25±0.26)粒/m2;(3)天山云杉林土壤种子库年际间物种组成的相似性不高,Czekanowski相似系数均值仅为0.344,并随间隔年限的增加呈现减小—增大—减小的变化趋势.天山云杉林土壤种子库物种组成和种子密度稳定性差,年际间相差悬殊,物种组成的相似性不高,种子库中天山云杉种子密度主要受其种子库采样前一年天山云杉结实丰歉的影响,属间断型.土壤种子库年际变化特征可为天山森林的更新恢复和可持续经营提供科学依据.     

Disturbance regimes as determinants of seed banks in coastal dune vegetation of the southeastern Cape

Soil-stored seed banks of grassland, fynbos and thicket, all growing on calcareous dunes and each subject to different disturbance regimes, were examined. Seed banks were determined from counts of germinants from 50 soil cores from each type. Aboveground estimates of plant species cover in 10 1-m² plots were used in determining vegetation/seed bank similarities. There was no evidence for seed bank densities to be markedly higher in the most frequently disturbed community (grassland -4273 seeds/m²) than the least disturbed community (thicket - 3417 seeds/m²). Highest similarity between seed bank and above-ground vegetation composition in terms of species and growth form/life-span classes was recorded for grassland (CC = 50%). Lowest similarity (CC = 13%) was found in the less frequently disturbed thicket where no seeds of climax trees were recorded in the seed bank. A fynbos community on a north-facing (warm, dry) slope had intermediate-sized seed banks (1683 seeds/m²) with intermediate vegetation/seed bank similarity (CC = 46%). However, on the south-facing slope, which has a large post-fire ephemeral herb component, seed banks were larger (4518 seeds/m²) but less similar to above-ground vegetation (CC = 39%o). Ordination (DCA) of vegetation data from the four communities was different from an ordination of their seed bank data. Fynbos shrub species were absent from seed banks of both grassland and thicket, even though secondary succession proceeds from grassland, through fynbos to thicket. Their seed banks appear less persistent than those of European heath or Californian chaparral shrubs.     

The dynamics of the soil seed bank after a fire event in a woody savanna in central Brazil

The seed-bank dynamics of cerrado, a savanna-like vegetation type in central Brazil, was monitored for a year after a fire event in the mid-dry season. Fifty paired soil and litter samples were collected 1 day before and 1 day after the fire to record the immediate effects on the seed bank, and thereafter at monthly intervals to investigate the post-fire seed bank dynamics. The samples were hand-sorted and the intact seeds were classified as monocot or dicot and counted. All seeds underwent germination trials in a germination chamber for 1 month. Seeds that did not germinate were checked for the presence and viability of the embryo. The sorted soil samples were placed in a greenhouse for 6 months, and the count of emerging seedlings was added to the number of germinated and dormant seeds from the germination trials to estimate the total number of viable seeds per sample. The fire did not affect the total seed-bank density: 63 ± 8 seeds m^?2 before the fire, and 83 ± 20 seeds m^?2 (mean ± se) immediately after it. Although monocots represented 65 % of the pre-fire seed bank, 1 year after the fire, the monocot seed density did not reach the pre-fire value, whereas the density of dicot seeds increased threefold. After the fire, the viable seed density and species richness, decreased with the onset of the rainy season coinciding with germination in the field. Therefore, post-fire recruitment increases genetic variability and contributes to the persistence of plant populations in cerrado communities.     

Germinable soil seed banks in a tropical savanna: seasonal dynamics and effects of fire.

Abstract The germinable soil seed bank of a tropical eucalypt savanna of north‐eastern Australia was found to be dominated by grasses and forbs, with seed bank density ranging from 58 to 792 seeds per square metre, from a total of 53 species. Late dry season fires and the fire‐related cues, heat shock and smoke, broke the seed dormancy of a range of tropical savanna species. Heat shock promoted the germination of the species groups natives, exotics, subshrubs, ephemeral and twining perennial forbs, and the common species Indigofera hirsuta, Pycnospora lutescens and Triumfetta rhomboidea. Exposure to smoke at ambient temperature promoted germination from the soil seed bank of the species groups combined natives, upright perennial forbs and grasses, as well as the common grasses Digitaria breviglumis and Heteropogon triticeus. The germinable soil seed bank varied seasonally, increasing from the mid wet season (February) and early dry season (May) to a maximum in the late dry season (October). The effect of recent fire history on soil seed bank dynamics was limited to the immediate release of some seed from dormancy; a reduction in seed densities of subshrubs and monocots, other than grasses, in recently burnt savanna; and enhanced seed density of the ephemeral I. hirsuta in the year following fire. The seed banks of most savanna species were replenished in the year following burning.     

Smoke and heat: influence on seedling emergence from the germinable soil seed bank of mesic grassland in South Africa

We examined if germination and seedling emergence of species from the soil seed bank of mesic grassland in South Africa differed in their response to smoke or heat treatments alone or combined. Soil seed bank samples taken from 0 to 5 cm depth of the topsoil were treated with smoke-water solution, heat and heat + smoke-water combined and subsequent emergence of seedlings monitored over 90 days. In total, 790 individuals from 11 different plant families representing 37 graminoid (Poaceae) and non-graminoid species of flowering plants were identified. The most abundant plant families that emerged were Poaceae followed by Asteraceae. Grasses contributed 18% of the germinable seed bank, while non-grass (forbs and trees) species contributed >80%. The most abundant grass species that emerged was Themeda triandra and the most common forb species was Centella asiatica. Compared to the control, smoke-water (SW) and/or smoke-water + heat (H + SW) treatment resulted in a 1.6-fold increase in mean number of seedling emergence. SW and H + SW treatment of soil seed bank also resulted in fourfold to fivefold greater biomass production. These results suggest that emergence of some smoke-responsive perennial grasses (e.g. Themeda triandra) from the soil seed bank can be enhanced using smoke techniques. Moreover, plant-derived smoke could potentially have a significant role in the restoration of degraded grasslands in South Africa and elsewhere.     

Effects of elevated nitrogen and exotic plant invasion on soil seed bank composition in Joshua Tree National Park

We investigated the effects of exotic species invasion and 3?years of nitrogen (N) fertilization on the soil seed bank in Joshua Tree National Park, California, USA at four sites along an N deposition gradient. We compared seed bank composition and density in control (no N added) and fertilized (30?kg?N?ha^?1?year^?1) plots to determine if the seed bank would reflect aboveground changes due to N fertilization. Soil samples were collected and germinated in a greenhouse over 2?years. In the field, invasive species cover responded positively to N fertilization. However, we did not observe increased seed density of exotic invasive species in fertilized plots. While no significant differences were detected between treatments within sites, exotic invasive grass seeds overwhelmed the seed bank at all sites. Significant differences between sites were found, which may be due to differences in level of invasion, historic N deposition, and soil surface roughness. Sites experiencing low N deposition had the highest seed bank species richness for both control and fertilized treatments. Aboveground plant density did not correlate well with seed bank density, possibly due to the inherent patchiness of soil seed banks and differential ability of species to form seed banks. This seed bank study provided insight into site-specific impacts on native versus invasive species composition of soil seed banks, as well as magnitude of invasion and restoration potential at invaded sites.     

The seed bank in a subtropical freshwater marsh: implications for wetland restoration

Seed bank samples were collected from Huli Marsh, a subtropical shallow water mountainous marsh in Hunan Province, South China. Core samples were divided into upper and lower layers (each 5 cm in depth) and allowed to germinate in three water levels (0, 5 and 10 cm) over a 4-month period. A total of 51 species germinated and the mean density was 9211 ± 7188 seedlings m⁻². In the top 5 cm 41 species and 5747 ± 5111 seedlings m⁻² germinated, whereas 40 species and 3464 ± 3363 seedlings m⁻² did so from 5–10 cm. Germinated seedling density was significantly higher in the upper layer, largely due to differences in eight species. With increasing experimental water depth, less seedlings germinated: respectively, 9788 ± 7157 m⁻², 2050 ± 2412 m⁻² and 1978 ± 2616 m⁻², of 44, 21 and 19 species, submerged under 0, 5 or 10 cm. Seven species could emerge only in 0 water level. Vallisneria natans occurred only in 5 cm water, whereas Ottelia alismoides occurred in 10 cm water. In the vegetation survey of the marsh, 25 species were recorded, which was less than half of the species recorded in the seed bank. The top 10 dominants in the standing vegetation, accounting for 89% of vegetation abundance, represented only 10% in the seed bank. Twenty germinated species that also occurred in the standing vegetation accounted for 56% of the total seed bank. Our observed number of species germinating from a Chinese wetland seed bank is within the range observed elsewhere in the northern hemisphere (15–113 species).     

Soil seed bank and regeneration potential on eroded hill slopes in the Kondoa Irangi Hills,central Tanzania

Abstract. The soil seed bank of the severely eroded Kondoa Irangi Hills, Tanzania was studied in order to determine the seed density and composition and to establish the relationship between seed bank and standing vegetation. The area had not been grazed for 15 yr prior to the study. A recently grazed area was used for comparison. The seed bank density (at 0 - 5 cm depth) ranged from 344 to 915 8 seeds/m² in the dry season and 172 to 5107 seeds/m² in the wet season. The seed bank was very heterogeneous, both spatially and temporally, and it showed significant variation in size and composition in both sampling periods. The species similarity between the seed bank and the above-ground vegetation in all plots was low (Sørensen's index = 0.00–0.44). The highest similarity was found in the recently grazed area. The seed bank was dominated by annuals and by early successional species. It is concluded that re-vegetating the hill slopes with woody vegetation by using the seed bank will be difficult because seeds of woody species were not found in the soil.     

Soil seed banks in church forests of northern Ethiopia: Implications for the conservation of woody plants

Church forests are sanctuaries for different organisms, ranging from microbes to large animals, which have almost disappeared in most parts of northern Ethiopia. Despite the actual and potential significance of these forests, studies and documented information on their bio-physical features and socio-economic setting are either scanty or totally lacking. A study was, therefore, carried out in seven of these church forests, namely Hiruy, Zahara, Gibtsawit, Gelawdiwos, Dengolt, Debresena and Ascha with the objective of assessing the composition, densities and spatial heterogeneity of soil seed banks. The number of viable seeds in the soil samples, from both germination and sieving, corresponded to a seed bank density down to 9 cm in the soil of 7594 seeds m⁻² at Ascha, 2064 seeds m⁻² at Debresena, 4208 seeds m⁻² at Dengolt, 3158 seeds m⁻² at Gelawdiwos, 2754 seeds m⁻² at Gibtsawit, 2759 seeds m⁻² at Hiruy, and 1909 seeds m⁻² at Zahra. The total number of species recorded was 50, representing at least 22 plant families, in the seven church forests, of which 13 were recorded from Ascha, 29 from Debresena, 26 from Dengolt, 19 from Gibtsawit, 22 from Gelawdiwos, 20 from Hiruy and 21 from Zahra. There was a significant difference in seed densities and number of species recovered from the soil seed banks of the seven church forests. Ascha had a significantly higher seed bank density than all other church forests. Dengolt exhibited a significantly higher number of species compared with other church forests. Of the identified species recorded from the seven church forests, 83% were herbs, 11% trees and 6% shrubs. Our results revealed that church forests accumulate large quantities of persistent seeds of herbaceous species in the soil while only five (6%) of the 91 woody species recorded in the standing vegetation of the seven church forests were represented in the soil seed banks. The fact that most of the dominant tree species do not accumulate seeds in the soil suggests that their regeneration from seeds would be prevented by removal of mature individuals in the standing vegetation. The results also provided further evidences that consolidate the conclusions of previous studies, i.e. the future existence of the woody flora characteristic of dry Afromontane areas in Ethiopia depends on the conservation and sustainable utilization of the few remnant natural forests.     

Soil Seedbank in a Dipterocarp Rain Forest in Xishuangbanna,Southwest China1

Dipterocarp rain forest reaches its northern latitudinal limit in Xishuangbanna, Southwest China. We studied the soil seedbank of dipterocarp rain forest in Xishuangbanna during the dry and wet seasons. Results showed that there were large seed accumulations in both the dry (mean ± SD; 3925 ± 2533 seeds/m²) and wet seasons (5415 ± 3232 seeds/m²). One hundred and sixteen species of seed plants were identified from germination, 66 percent of which were woody species. Weed or pioneer species dominated the seedbank. The soil seedbank in Xishuangbanna had similar species composition as compared to those in tropical Asia, but higher seed storage reflects the intense disturbance and forest fragmentation in this area.     

Red crabs in rainforest on Christmas Island,Indian Ocean: No effect on the soil seed bank

Red land crabs (Gecarcoidea natalis) are both abundant and widespread in intact rainforest on Christmas Island, Indian Ocean, occurring at densities of ~ 1 crab m^–2. We used an on-going exclusion experiment to test the hypothesis that through their activities as seed predators and soil disturbers, these omnivores are important determinants of the density and relative species abundance of seeds in the soil seed bank of undisturbed primary rainforest. After 6.5 y of continuous crab exclusion, there was no significant difference in the density of seeds in the top 3 cm of soil between the control (3671 ± 517 seeds/0.19 m²) and exclusion plots (3285 ± 801 seeds/0.19 m²), nor was there any significant treatment difference in relative species composition. There were also no significant treatment differences when Muntingia calabura, which accounted for 91.9% of all seeds in the seed bank, was excluded from the analyses. We attribute the lack of a significant impact of red crabs to their probable inability to handle the very small seeds which occurred in the seed bank, and the low level of soil disturbance at the study site. We conclude that the observed impact of red crabs on the abundance and diversity of successional species in natural light gaps on Christmas Island is due to their impact on germinating seedlings alone, rather than any additional impact on the soil seed bank.