Seed bank dynamics of two obligate seeders, <Emphasis Type="Italic">Cistus monspeliensis</Emphasis> and <Emphasis Type="Italic">Rosmarinus officinalis</Emphasis>, in relation to time since fire

Many species in Mediterranean-type ecosystems regenerate after fire by seed germination from soil seed banks. Seed bank dynamics of two of those obligate seeders, Cistus monspeliensis and Rosmarinus officinalis, were investigated in relation to stand age since fire in southwestern Portugal. Soil seed density, annual seed input, annual seed losses through germination and seed persistence were compared between species at stands differing in age since fire (5, 10 and 35 years). Soil seed density and seed input increased over the first decade after fire and were lowest at 35-year-old stands for C. monspeliensis. In R. officinalis, few seeds were produced and found in the soil at early stages, and maximum seed input and soil seed density were attained at 35-year-old stands. Soil seed density was mostly driven by seed production in both species, which is largely dependent on plant traits and population dynamics related to fire. Overall, stand age since fire had a negligible effect on seed germination, seed persistence and viability. Ten to 39% of buried seeds were not recovered after 1 year, and viability of seeds recovered was 97–100% for C. monspeliensis and only 0–3% for R. officinalis. Variation in plant traits within the seeder syndrome was evidenced by this study. R. officinalis evidenced lower seed persistence, lower proportion of viable seed produced and lower density of viable soil seed than C. monspeliensis at any stage after fire. R. officinalis is expected to depend largely on previous year seed production for population replacement after fire.     

Rodent seed predation and seedling recruitment in mesic grassland

Seedling recruitment of two grasses (Arrhenatherum elatius and Festuca rubra) and two herbs (Centaurea nigra and Rumex acetosa) was measured in areas with and without rodents to which seeds of each species were sown at three seed densities (1000, 10,000 and 50,000 seeds m⁻²) in two seasons (spring and autumn 1995). Seed removal was measured for 10-day periods and the fate of seedlings was followed for 15 months after sowing. The proportion of seed removed ranged from 6 to 85% and increased with increasing seed density for each species. Rodents had no effect on seedling emergence or survival in the spring sowing. In the autumn sowing, rodents reduced seedling emergence of all four species sown at 1000 and 10,000 seeds m⁻² but had no impact at 50,000 seeds m⁻², presumably because of microsite limitation. We suggest the difference between spring and autumn arose because emergence was seed limited in autumn but microsite limited in spring; microsite availability was higher in autumn because a summer drought killed plants, reduced plant biomass and opened up the sward. Fifteen months after the autumn sowing, fewer A. elatius and C. nigra seedlings survived on plots exposed to rodents. This result reflected not only the reduced seedling emergence but also increased seedling mortality (seedling herbivory) in sites exposed to rodents. In contrast, F. rubra and R.acteosa showed density-dependent seedling survival which compensated for initial differences in seedling emergence, so that no effect of rodents remained after 15 months. The results suggest that rodent seed predation and seedling herbivory exert strong effects on seedling recruitment of A.elatius and C. nigra when recruitment conditions are favourable (conditions that lead to high microsite availability) and may contribute to both species being maintained at low densities in the grassland. The results also demonstrate that highly significant impacts of rodent seed predation at the seedling emergence stage can disappear by the time of plant maturation. Received: 2 March 1998 / Accepted: 28 September 1998     

Dispersal and recruitment dynamics in the fleshy‐fruited Persoonia lanceolata (Proteaceae)

Question: What is the role of dispersal, persistent soil seed banks and seedling recruitment in population persistence of fleshy‐fruited obligate seeding plant species in fire‐prone habitats? Location: Southeastern Australia. Methods: We used a long‐term study of a shrubby, fleshy‐fruited Persoonia species (Proteaceae) to examine (1) seed removal from beneath the canopy of adult plants; (2) seedling recruitment after fire; (3) the magnitude and location of the residual soil seed bank; and (4) the implications for fire management of obligate seeding species. We used demographic sampling techniques combined with Generalised Linear Modelling and regression to quantify population changes over time. Results: Most of the mature fruits (90%) on the ground below the canopy of plants were removed by Wallabia bicolor (Swamp wallaby) with 88% of seeds extracted from W. bicolor scats viable and dormant. Wallabies play an important role in moving seeds away from parent plants. Their role in occasional long distance dispersal events remains unknown. We detected almost no seed predation in situ under canopies (< 1%). Seedling recruitment was cued to fire, with post‐fire seedling densities 6‐7 times pre‐fire adult densities. After fire, a residual soil seed bank was present, as many seeds (77‐100%) remained dormant and viable at a soil depth where successful future seedling emergence is possible (0‐5 cm). Seedling survival was high (> 80%) with most mortality within 2 years of emergence. Plant growth averaged 17 cm per year. The primary juvenile period of plants was 7–8 years, within the period of likely return fire intervals in the study area. We predicted that the study population increased some five‐fold after the wildfire at the site. Conclusions: Residual soil seed banks are important, especially in species with long primary juvenile periods, to buffer the populations against the impact of a second fire occurring before the seed bank is replenished.     

Ant-mediated expansion of an obligate seeder species during the first years after fire

Most obligate seeder species build up a soil seed bank that is associated with massive seed germination in the year immediately after a fire. These species are also shade‐intolerant and disappear when vegetation cover closes, creating unsuitable conditions for seedling recruitment. The only way for these plants to expand their populations is when habitats suitable for seedling recruitment arise (i.e. in years immediately after a fire). However, short primary seed dispersal of obligate seeders does not allow these plants to colonise the suitable habitats, and these habitats can only be colonised by secondary seed dispersion. We hypothesised that Fumana ericoides, an obligate‐seeding small shrub, not only establishes abundantly in the first year after fire, but also expands its local range in the following years due to secondary dispersal by ants while suitable habitats are still available. We tested this hypothesis using experimental studies and a simulation model of potential population expansion in a recently burned area. Results showed that F. ericoides not only established prolifically in the year immediately after fire, but was also able to recruit new individuals and expand its population in the years following the fire, despite a low germination rate and short primary seed dispersal. Ant‐mediated seed dispersal and availability of suitable habitats were key factors in this phenomenon: ants redistributed seeds in suitable habitats while they were available, which accelerated the expansion of F. ericoides because new plants established far away from the core population.     

Seed predation limits post-fire recruitment in the waratah (<Emphasis Type="Italic">Telopea speciosissima</Emphasis>)

Seed predation may reduce recruitment in populations that are limited by the availability of seeds rather than microsites. Fires increase the availability of both seeds and microsites, but in plants that lack a soil- or canopy-stored seed bank, post-fire recruitment is often delayed compared to the majority of species. Pyrogenic flowering species, such as Telopea speciosissima, release their non-dormant seeds more than 1 year after fire, by which time seed predation and the availability of microsites may differ from that experienced by plants recruiting soon after fire. I assessed the role of post-dispersal seed predation in limiting seedling establishment after fire in T. speciosissima, in southeastern Australia. Using a seed-planting experiment, I manipulated vertebrate access to seeds and the combined cover of litter and vegetation within experimental microsites in the 2 years of natural seed fall after a fire. Losses to vertebrate and invertebrate seed predators were rapid and substantial, with 50% of seeds consumed after 2 months in exposed locations and after 5 months when vertebrates were excluded. After 7 months, only 6% of seeds or seedlings survived, even where vertebrates were excluded. Removing litter and vegetation increased the likelihood of seed predation by vertebrates, but had little influence on losses due to invertebrates. Microsites with high-density vegetation and litter cover were more likely to have seed survival or germination than microsites with low-density cover. Recruitment in pyrogenic flowering species may depend upon the release of seeds into locations where dense cover may allow them to escape from vertebrate predators. Even here, conditions suitable for germination must occur soon after seed release for seeds to escape from invertebrate predators. Seed production will also affect recruitment after any one fire, while the ability of some juvenile and most adult plants to resprout after fire buffers populations against rapid declines when there is little successful recruitment.     

Establishment of microscale vegetation pattern in maritime chaparral after fire

We quantified microscale pattern in vegetation and seed assemblages along a 24 m transect before and for two years following a controlled burn in chamise (Adenostoma fasciculatum) chaparral in central coastal California. Our objective was to document scale-dependent correlation between pre-burn seed assemblages, pre-burn canopy cover, microtopography, soil temperatures during burning, post-burn seed assemblages and post-burn vegetation. Scale-dependent correlations among pre- and post-burn seed densities, maximum soil temperatures during burning, microtopography and post-burn vegetation were measured based on two-term local covariance analysis.Seed distribution varied among species prior to fire, with seeds of some annual species concentrated in gap areas. Maximum soil temperatures during burning ranged from less than 50 °C to 225 °C, and were generally lowest in gaps in the pre-burn canopy. These gaps were associated with local topographic depressions. After burning, readily germinable seeds were concentrated in or near gaps in the pre-burn canopy. Germination of different species was variously enhanced, diminished or unchanged by the passage of fire. Post-burn vegetation was very patchy, with some areas nearly devoid of seedlings and other areas, especially pre-burn canopy gaps, supporting numerous seedlings. Seedling recruitment patterns in the second year were generally highly correlated with patterns in pre-burn seed banks and first year vegetation. Although many species exhibited similar recruitment patterns, several different mechanisms may have been responsible for the origin of those patterns.     

The effects of drought and fire on seed and seedling dynamics in a tropical seasonal forest in Thailand

Seed production and seedling dynamics were studied for 6 years(1992–1997) with particular emphasis on the effects of the drought andfire in a natural tropical seasonal forest at Mae Klong Watershed ResearchStation, Kantchanaburi, western Thailand. The number of seed produced andemerged seedlings varied greatly among years and species. Most tree speciesproduced very few seeds in 1994 and many in 1995 and 1996, particularly,Shorea siamensis Miq. which displayed the most seedinghabit. Some seedlings, Shorea, Dipterocarpus alatus Roxb.,and Dipterocarpus turbinatus Gaertn. f., emerged at theendof the dry season (late April, after the first rain but when soils still have alow soil moisture content), while the majority of seedlings emerged at thebeginning of the rainy season. After a fire disturbed the plot in April 1996,many species increased their seedling emergences, especially Berryaammonilla. Seedlings of most species less than a year old showed arelatively low mortality in the rainy season, but those ofPterocarpusmacrocarpus Kurz had high mortality. The mean survival rate for allseedling species in this forest was quite low (24.1% y^–1), anddifferent values existed for the rainy and the dry season (11.5, and 6.1%month^–1, respectively). This suggests that different specieshave adaptations related to the season of seedling emergence and resistance todrought in this tropical seasonal forest community. The traits of tree speciesare classified in terms of their traits in the early stages of regeneration;seed and/or seedling bank, resprouting, and drought resistance. Most specieshave adapted to fire and/or drought by resprouting, seed bank, and/or seedlingbank, although the few species which occur mainly in mesic evergreen forestshave less adapted to these environments. The demographic variations in seed andseedling stages may contribute to the coexistence of these species in thisspecies-rich dry tropical forest.     

Aggregated seed dispersal by wreathed hornbills at a roost site in a moist evergreen forest of Thailand

Hornbills (Bucerotidae) are widely regarded as important seed dispersers in tropical forests in Africa and Asia. We investigated how the roosting behavior of wreathed hornbills (Aceros undulatus) influences seed deposition and seedling survival at a roost site in a moist evergreen forest of Khao Yai National Park, Thailand. Fallen fruits and seeds were collected in traps that were placed around a roosting site for 14 months, and seedlings were monitored in adjacent quadrats for 3 years. Seedfall and seedlings of species represented in the hornbill diet occurred at significantly higher densities in the traps and quadrats located beneath the crown of the roosting tree than in those located beyond the crown. With the exception of Cinnamomum subavenium, the seeds and seedlings of most diet species rarely survived beyond the first year. The quality of hornbill dispersal to this roosting site may be poor due to the highly concentrated seedfall, which results in high seed and seedling mortality. However, the number of seeds deposited by each hornbill each day at roosting sites is relatively low. Wreathed hornbills are primarily scatter dispersers during the day and probably serve as agents of seed dispersal in the moist evergreen forest of Khao Yai.     

Seed banks of Erica australis and Calluna vulgaris in a heathland subjected to experimental fire

Abstract. Soil samples were collected before and after an experimental fire on a heathland in the province of León (Spain). The seed banks were assessed by counting the numbers of seedlings appearing of Erica australis and Calluna vulgaris. A total of 2285 germinable seeds/m² before burning and 1177 germinable seeds/m² after the fire were estimated for Erica australis, and 90 and 690 germinable seeds/m², respectively, for Calluna vulgaris. After the fire the number of germinable seeds/m² of Erica australis had decreased, whilst there was a considerable increase in the number of germinable seeds/m² of Calluna vulgaris. Neither seedlings nor sprouts of Calluna vulgaris had appeared in the field plots 10 months after the fire. Erica australis did recover mainly by sprouting after fire.     

Soil seed densities and emergence patterns in pastures in the seasonally dry tropics of northeastern Australia

Germinable seed densities in the surface (0–10 cm) soil of pasture communities growing at Lansdown, near Townsville, were measured during the late dry season before the first germinating rain and again during the following wet season after germination but before the input of new seed. Seedlings emerging in the field were counted at approximately weekly intervals during this period to determine emergence patterns. Twelve communities were sampled in 1980–81 and six were re-sampled in 1981–82. During the late dry season germinable seed densities ranged from 5000 to 40 000 seeds m^-2. Seeds of the introduced legume, Stylosanthes hamata, were present in all pastures. There were many seeds of annual grasses (Digitaria ciliaris and Brachiaria miliiformis) and sedges (Cyperus and Fimbristylis spp.) but only few seeds of perennial grasses (both native and introduced). Soil seed densities were much lower during the wet season than during the preceding dry season, particularly for the grasses. Emergence commenced and approximately 70% of all seedlings emerged on the first major rainfall of the wet season. The subsequent emergence pattern varied between years. In 1980–81 there was a gradual and continuous increase in seedling numbers under the continuously moist conditions which prevailed. In 1981–82 further emergence occurred in discrete events related to rainfall and intervening dry periods. Maximum seedling densities exceeded 34 000 seedlings m^-2 including 29 000 grass seedlings (mainly annual species). The implications of these results for species survival and pasture composition are discussed.     

Recovery patterns of three chaparral shrub species after wildfire

Summary In a mature, even aged stand of mixed chaparral, Rhus laurina (facultative resprouter) had consistently higher water potentials and deeper roots than Ceanothus spinosus (facultative resprouter) and Ceanothus megacarpus (obligate seeder). For two years following a wildfire, the same stand of chaparral had resprouts with higher survivorships, predawn water potentials, stomatal conductances, photosynthetic rates and shoot elongation rates than seedlings. Supplemental irrigation of seedlings during summer months removed differences between resprouts and seedlings suggesting that the cause of such differences was limited water availability to the shoot tissues of seedlings. After two years of postfire regrowth, mean seedling survivorship for the obligate seeder (C. megacarpus) was 42%, whereas seedling survivorship for facultative resprouters was only 18% (C. spinosus) and 0.01% (R. laurina). Our results are consistent with the hypothesis that lack of resprouting ability among obligate seeders is offset by an enhanced ability to establish seedlings after wildfire, allowing obligate seeders to maintain themselves in mixed populations through many fire cycles.     

The influence of dominant shrubs,fire, and time since fire on soil seed banks in mixed chaparral

The composition and density of soil seed banks beneath co-occurring Adenostoma fasciculatum and Ceanothus greggii shrubs from three chaparral stands last burned 9, 35 and 85 years before 1986 were investigated. The overall density of seeds in the soil, as estimated by germinations under greenhouse conditions, increased with time since fire (ca. 8000 to 25000/m²). However, this increase was due entirely to the accumulation of A. fasciculatum seed in the soil (ca. 2000 to 21000/m²). In contrast, the density of C. greggii seed was different in each of the three stands, but was not correlated with time since fire: maximum densities were recorded from the 35 year old stand (ca. 2000/m²).A total of 31 taxa germinated and 17 occurred in sufficient numbers to be analyzed statistically. Germinable seed densities of three herb species were not influenced by soil source (beneath A. fasciculatum or C. greggii), time since fire, or the direct effects of a controlled fire treatment. Germinable seed densities of a further nine species were significantly influenced by the elapsed time since stands last burned. The densities of four decreased and five increased. Four of the species that increased in seed density over the three stands were annuals, suggesting that the chaparral sub-canopy habitat is not as unfavorable for annuals as is often assumed. The fire treatment decreased germinable seed densities of four annual species by 40–70%, but increased the germinable seed densities of the shrubs A. fasciculatum and C. greggii, and the annual Phacelia brachyloba. Our results indicate that seeds of A. fasciculatum will increase in the soil bank for at least 85 years after fire in chaparral where it is dominant. In contrast, seed reserves of C. greggii appear to be influenced primarily by site-specific patterns of seed production and by the intensity of post-dispersal seed predation.     

Sabal palmetto seed size: causes of variation,choices of predators,and consequences for seedlings

High variation in seed size, as is common among angiosperms, may be maintained in a plant species when several factors select for seed size. Variation may also result from differences among adult plants, such as nutrient and water availability or the amount of photosynthetic tissue. In a study of Sabal palmetto seed ecology I found high seed size variation both within- and among-palms, and investigated possible factors maintaining this variation. Seed size was positively correlated with the number of leaves on parent palms. Larger seeds produced more vigorous seedlings that had greater leaf length, area, and mass, and greater root mass. Caryobruchus gleditsiae (Bruchidae: Coleoptera), whose larvae develop within palm seeds, preferentially oviposited on larger seeds, which in turn produced larger beetle offspring. By choosing the largest seeds available, ovipositing beetles thus affect both the quantity and the quality of seeds available for recruitment. I conclude that because beetle predation selects against large seeds, while larger seeds promote seedling vigor, the maintenance of seed size variation may be an adaptation of S. palmetto promoting both seed escape from predators and seedling vigor.     

Delayed post-fire seedling emergence linked to season: a case study with Leucopogon species (Epacridaceae)

In many fire-prone habitats around the world, natural fire regimes have shaped the evolution of the associated flora. Critical life history stages are often linked to fire in species that occur in these fire-prone habitats but many species are unstudied. We investigated seedling emergence patterns over time, after several fires in south-eastern Australia, for three obligate seeders in the Genus Leucopogon (Family Epacridaceae): L. setiger, L. esquamatus and L. exolasius. Fixed quadrats were monitored for 12 to 30 months for newly emerged seedlings, both after fire and in unburnt L. exolasius and L. esquamatus habitats. There was a flush of seedling emergence in the first year after fire for all three Leucopogon species, with smaller pulses recorded in subsequent years. Time elapsed between fire and the onset of emergence differed between fires, but not between Leucopogon species. Whatever the timing of the fire, seedling emergence was restricted to the late autumn and winter periods, coinciding with emergence in unburnt habitat. This contrasts with patterns of emergence previously reported for other taxa, and also in this study, where emergence of the Leucopogon species after fire was delayed compared with co-occurring species in other genera. Our results suggest that seasonal factors are important to the germination ecology of these species and that combinations of fire-related and seasonal factors are necessary to maximise germination. Rainfall has a non-seasonal pattern in the study region and seasonal emergence post-fire has not previously been recorded. The magnitude of delay to emergence of species with seasonal emergence patterns will be determined by the season of fire but not by a seasonal pattern of rainfall in the post-fire year. A shift of the peak fire season could increase this delay, possibly affecting plant population recovery. This revised version was published online in June 2006 with corrections to the Cover Date.     

A test of the escape and colonization hypotheses for zoochorous tree species in a Western Amazonian forest

In order to assess the importance of seed dispersal (escape and colonization hypotheses), I used transplant experiments for seeds and seedlings of 5–11 plant species with fleshy fruits in a lowland tropical forest (Tinigua National Park, Colombia). I controlled seed density, distance to parental tree, and habitat type. I monitored seed removal, seedling survival, and seedling growth during the first year of development for an average of 554 seeds and 169 seedlings for each species. I supplemented the experimental results with measurements of natural recruitment. I found little support for the escape hypothesis during the seed and seedling stages. For six species that showed differences in seed removal associated with distance, five showed highest removal away from, than close to parent trees, suggesting predator satiation. Seedling survival during the first year was not consistently associated with low densities and long distances from parent trees. For the majority of species, seedlings did not survive flooding in low basins, and there was growth advantage for most plant species in canopy gaps. These differences imply advantages for seed dispersal to adequate habitats, as predicted by the colonization hypothesis. In contrast to experiments, strong negative distance-dependent effects were evident when analyzing natural recruitment patterns. The ratio between saplings and seedlings was higher away from parent trees for the species with enough recruitment to be analyzed and this suggests that a negative distance-dependent effect may also occur after seedling establishment. This pattern is suspected for several other species, but an analysis with some of the other most common trees showed a variety of negative, neutral, and positive distance dependent effects. This study emphasizes the importance of long-term studies to asses the role of seed dispersal.     

Establishment and reproduction of Aeschynomene virginica (L.) (Fabaceae) a rare,annual, wetland species in relation to vegetation removal and water level

Aeschynomene virginica is a rare annual plant found in freshwater tidal wetlands of the eastern United States. We hypothesized that standing vegetation and water inundation were two important environmental factors in its population dynamics. To test these hypotheses, we sowed seeds into plots with undisturbed vegetation or plots with all aboveground vegetation removed in 1998 and 1999. Presence/absence of seedlings was noted and seedling survival to reproduction, final size, and seed set were measured throughout both growing seasons. Seedling establishment from germination to the first true leaf stage increased with decreasing water depth. Vegetation removal plots had greater seedling establishment, higher seedling survival, and higher seed set per plant than non-removal plots. In a greenhouse study designed to test the effects of water level on seed germination and seedling establishment, no seedlings established in submerged soils, and seed germination and seedling establishment were lower in waterlogged soil than in wet soil. Physical stress associated with deeper water likely limits the distribution of A. virginica to higher elevations, where seeds that colonize patches with low vegetative cover are more likely to produce reproductive adults that produce more seeds relative to patches with established vegetation. A. virginica appears to be a fugitive species specializing on open habitat patches in tidal wetlands. This species may be dependent on disturbances for population establishment and maintenance.     

Aspects of variation in a neotropical seed dispersal system

This paper explores the causes and consequences of seed and seedling mortality of the tree Virola nobilis (Myristicaceae) Central Panama in order to understand the advantage to local seed dispersal by birds and monkeys. Post-dispersal mortality due to insects (primarily Conotrachelus spp., Curculionidae) accounts for 30–35% of seed and seedling death during the first 12 weeks after seed fall. Because more seeds and seedlings are killed under and near fruiting trees than 15–45 m away, seed dispersal confers a 20–40 fold advantage on seeds carried 45 m from fruiting adults. In contrast, >60% of seed and seedling death during the first year is due to seed predation by mammals, with >90% due to mammals among the <2% seeds that survive until maternal endosperm is exhausted ±12 weeks after seed fall. Mortaliy due to mammals is independent of distance from parent trees, confering no advantage to seed dispersal. Insects account for variation in mortality attributable to distance effects, mammals to between site effects.Early weevil infestations put a premium on seed removal by large birds (Ramphastos swainsonii, R. sulfuratus, Penelope purpurascens), which carry >50% of the seeds that they eat >40 m, as compared with smaller birds (Baryphthengus martii, Tityra semifasciata, Trogon massena) and monkeys (Ateles geoffroyi), which leave most or all of the seeds that they eat under or near the tree crown.     

Organisation of the soil seed bank in mixed chaparral

Seed densities, estimated from germinations under glasshouse conditions, were determined for fire-treated and control soils sampled from north and south facing stands of mixedAdenostoma fasciculatum-Ceanothus greggii chaparral that last burned 10, 17, 36, 62 and 86 years before 1987. A total of 53 species was recorded. Total germinable seed density was not influenced by aspect nor by an experimental fire treatment, however, densities varied significantly, but not predictably, over the time since fire sequence. Species diversity was significantly lower in the 86 year old stand, but neither diversity or richness was influenced by the fire treatment. Soils from north facing slopes supported a higher species diversity overall, and seed densities of six species were significantly higher on north than on south slopes.We identify two functional groups in the seed bank on the basis of the direct effects of fire on seed densities: a fire-dependent group, comprising 6 species and about 21% of the total seed bank, in which germinable densities increased after the fire treatment, and a fire-independent group, made up of 11 species and almost two-thirds of the seed bank. The fire-independent group is further separated into two: a shrub-centred sub-group (Crassula erecta, Filago californica, Pterostegia drymarioides, Streptanthus heterophyllus andCryptantha intermedia) maintained larger seed banks in older stands and was associated with shrub cover; and an opportunistic sub-group (Camissonia hirtella, Gnaphalium chilense, Mimulus pilosus andSenecio vulgaris) tended to have larger seed banks in younger stands and was associated with canopy gaps.Overall, the soil seed assemblage from this mixed chaparral does not appear to be tightly coupled to fire-generated opportunities for recruitment. The reason for this is that the fire-dependent group represents only 6 species. In addition, among the fire-independent species, the shrub-centred group of 5 annuals made up almost 40% of total soil seed density. This group shows ecological and taxonomic affinities with desert annuals.     

Early postfire seed dispersal, seedling establishment and seedling mortality of Pinus coulteri (D. Don) in central coastal California, USA

Seedling recruitment in many highly serotinous populations of Pinus coulteri on California's central coast depends almost entirely on periodic, stand-replacing fire. Compared to serotinous pines of the Mediterranean Basin, little detailed information is available on the postfire demography of California closed-cone pines, including P. coulteri. In September 1996 a wildfire burned the 760-ha American Canyon Research Natural Area (RNA). Using aerial photography, we mapped burn severity of P. coulteri-chaparral woodlands and forests within the RNA. From May to September of 1997, we also quantified seedling establishment and mortality in relation to biophysical site characteristics including fire severity. Seventy-six percent of P. coulteri forests and woodlands experienced high-severity burns, 9% moderate-severity burns, and 15% low-severity or unburned. Of the 53 plots used for seedling counts, 70% were high-severity, 26% moderate-severity, and 4% low-severity. Seedling densities 13 months postfire were low (0.21 m^–2), but seedling mortality also was low (8.4%). Aerial seed bank size increased from north-facing to south-facing slopes and from high-severity to low-severity burns. Seedling recruitment was unrelated to burn severity and increased with the size of the canopy seed bank (cone density). Many seedlings established from rodent seed caches; 23% of the seedlings established in clumps from seeds cached by Dipodomys agilis, Chaetodipus californicus and Peromyscus maniculatus. Pinus coulteri seeds have low potential for dispersal by wind, but secondary dispersal by rodents moves seeds away from source trees and into neighboring chaparral. We discuss the potential importance of rodent seed caching to postfire demography of California and Mediterranean serotinous pines.     

Development of vegetation patterns in early primary succession

Question: We investigated colonisation filters in early plant community development on a glacial outwash plain. We asked if these were related to seed limitation or to a lack of safe sites, if topographical heterogeneity affected species patchiness and how species life cycles influence successional trajectories. Location: An outwash plain (Skeiðarársandur) in southeast Iceland. Methods: We identified surface heterogeneity at two different scales, ca. 10–15 cm (larger stones and established plants) and ca. 50 m (shallow depressions representing dry river beds) at two study sites. We quantified species cover, flowering plant density, seed production, seed rain, seed bank density, seedling emergence and seedling survival from June 2005 to June 2007 for the whole plant community, and measured seed production for five species. Results: Mean vegetation cover was <2.5% at the sites. Low emergence rates and high seedling mortality were the two main recruitment filters. Only 1.4% of seedlings emerging in 2005 survived into the 2007 growing season. Topographical heterogeneity had little effect on plant colonisation. High annual variation was recorded, and the two study sites (ca. 2 km apart) differed in their colonisation success. Of the five species, establishment of Cerastium alpinum and Silene uniflora was most limited by lack of seeds, whereas establishment of Luzula spicata, Poa glauca and Rumex acetosella was most limited by safe sites. Conclusions: We conclude that colonisation processes and patterns in early primary succession on Skeiðarársandur were largely influenced by stochastic factors.