Scale-dependent diversity effects of seed dispersal by a wild herbivore in fragmented grasslands

Dispersal limitation between habitat fragments is a known driver of landscape-scale biodiversity loss. In Europe, agricultural intensification during the twentieth century resulted in losses of both grassland habitat and traditional grassland seed dispersal vectors such as livestock. During the same period, populations of large wild herbivores have increased in the landscape. Usually studied in woodland ecosystems, these animals are found to disperse seeds from grasslands and other open habitats. We studied endozoochorous seed dispersal by roe deer (Capreolus capreolus) in fragmented grasslands and grassland remnants, comparing dispersed subcommunities of plant species to those in the established vegetation and the seed bank. A total of 652 seedlings of 67 species emerged from 219 samples of roe deer dung. This included many grassland species, and several local grassland specialists. Dispersal had potentially different effects on diversity at different spatial scales. Almost all sites received seeds of species not observed in the vegetation or seed bank at that site, suggesting that local diversity might not be dispersal limited. This pattern was less evident at the landscape scale, where fewer new species were introduced. Nonetheless, long-distance dispersal by large wild herbivores might still provide connectivity between fragmented habitats within a landscape in the areas in which they are active. Finally, as only a subset of the available species were found to disperse in space as well as time, the danger of future biodiversity loss might still exist in many isolated grassland habitats.     

Seed bank dynamics of the desert cactus Opuntia rastrera in two habitats from the Chihuahuan Desert

In desert environments the main input to the seed bank of many succulents is the seed rain through zoochory while high levels of granivory by rodents, birds and ants are the main cause of subsequent losses. In the patchy environment of arid lands the characteristics of both processes may vary between habitats causing differences in the recruitment of new genets. To test this hypothesis we used populations of the desert cactus Opuntia rastrera which has different recruitment rates in the two adjacent habitats where it grows. In Opuntia-dominated scrublands (nopaleras, density ca. 4,000 plants/ha) 1 seedling out of 7,000,000 seeds establish, whereas in grasslands (density ca. 100 plants/ha) this ratio is 1:20,000. From 1996 until 1998 the seed rain, seed removal by granivores and seed abundance in the soil were monitored in both habitats. Results showed striking differences in the dynamics of the seed bank of both habitats. Seed rain was 8.5 times bigger in nopaleras than in grasslands. In nopaleras most seeds were removed by rodents while the quantities of seeds removed by rodents, birds and ants in grasslands were similar. One year after dispersal (the time necessary to break seed dormancy) only 6% of original nopalera seeds and 12% of grassland seeds remained. After germination trials only 1% (ca. 15,000 seeds/ha) and 2% (ca. 2,500 seeds/ha) respectively were viable. These differences in the effective seed bank (6 times bigger in nopaleras) can not explain the differences in genet recruitment (which is several orders of magnitude bigger in grasslands). Apparently the between habitat difference in nurse plant availability and in rodent density (which inflict a strong hervibory upon seedlings) can explain the differences in genet recruitment. It is speculated that this between habitat difference in genet recruitment suggests that the species evolved in less extreme environments (e.g. grasslands) than desert scrublands which, in turn, are colonised due to the singular ability of O. rastrera for vegetative propagation. This revised version was published online in August 2006 with corrections to the Cover Date.     

Endozoochorous seed dispersal by free-ranging herbivores in an abandoned landscape

Seed dispersal is a crucial process for the dynamics and maintenance of plant populations. Free-ranging animals are effective dispersal vectors because they can move between similar habitats and transport seeds into favourable environments. Dung samples from two species of common free-ranging mammals—deer and wild boar—were used to study endozoochorous dispersal of seeds in a military training area in western Bohemia. The area was abandoned after WWII, and the military training area was established in 1953. The vegetation consists of shrublands and dry grasslands. Data on the local species pool of grassland herbs and forbs were collected to compare the characteristics of dispersed versus non-dispersed plants. Deer and wild boar dispersed 84 plant species; however, species composition of seedlings emerging from dung samples showed significant differences between dispersal vectors and notable change across the growing season. 80% of all seedlings extracted from the dung samples belonged to stinging nettle, Urtica dioica. From trait analyses, seeds of endozoochorous plants had a higher longevity index in the soil seed bank than non-endozoochorous plants and more often possessed a mucilaginous surface. Our results show that deer and boar are successful, though not substitutable dispersers.     

Secondary invasion of <Emphasis Type="Italic">Acer negundo</Emphasis>: the role of phenotypic responses versus local adaptation

During plant species invasions, the role of adaptive processes is particularly of interest in later stages of range expansion when populations start invading habitats that initially have not been disposed to invasions. The dioecious tree Acer negundo, primarily invasive in Europe in wet habitats along riversides and in floodplains, has increased its abundance in dry habitats of industrial wasteland and ruderal sites during the last decades in Eastern Germany. We chose 21 invasive populations from wet and from dry habitats in the region of Halle, Saxony-Anhalt, Germany, to test whether Acer negundo exhibits a shift in life-history strategy during expansion into more stressful habitats. We analyzed variables of habitat quality (pH, soil moisture, exchangeable cations, total C and N content) and determined density, sex ratio and regeneration of the populations. In addition, we conducted germination experiments and greenhouse studies with seedlings in four different soil moisture environments. Local adaptation was studied in a reciprocal transplant experiment. We found habitat type differentiation with lower nutrient and water supply at the dry sites than at the moist sites and significant differences in the number of seedlings in the field. In accordance, seeds from moist habitats responded significantly faster to germination treatments. In the transplant experiment, leaf life span was significantly larger for populations originating from dry habitat types than from moist habitats. This observed shift in life history strategy during secondary invasion of A. negundo from traits of establishment and rapid growth towards traits connected with persistence might be counteracted by high gene flow among populations of the different habitat types. However, prolonged leaf life span at dry sites contributed remarkably to the invasion of less favourable habitats, and, thus, is a first indication of ongoing adaptation.     

Recruitment requirements of the rare and threatened Juncus atratus

The long-term persistence of populations and species depends on the successful recruitment of individuals. The generative recruitment of plants may be limited by a lack of suitable germination and establishment conditions. Establishment limitation may especially be caused by the competitive effect of surrounding dense vegetation, which is believed to restrict the recruitment success of many plant species to small open patches (‘safe sites’). We conducted experiments to clarify the roles of germination and seedling establishment as limiting processes in the recruitment of Juncus atratus Krock., a rare and threatened herbaceous perennial river corridor plant in Central Europe. Light intensity had a positive effect on germination. However, some seedlings emerged even in total darkness and the germination rate at 1% light intensity was more than half of that at 60% light intensity. Seedling establishment in the field after 10 weeks was 30% on bare ground, but it was close to zero in grassland. Establishment in the growth chamber after 8 weeks was close to 75% for seedlings that germinated underwater, but only about 35% for seedlings that germinated afloat. Furthermore, establishment decreased with flooding duration on bare ground, but increased with flooding duration in grassland. These data indicate that establishment, rather than germination, is a critical life stage in Central European populations of J. atratus. They furthermore indicate that the competition of surrounding vegetation for water limits seedling establishment under field conditions without flooding, largely restricting establishment success to bare ground habitats. In contrast, grassland is more suitable for the recruitment of J. atratus than bare ground under prolonged flooding. Grassland may facilitate the establishment of J. atratus seedlings during long-lasting floods by supplying oxygen to the soil through aerenchyma. The shift from competition to facilitation in grassland occurred after 30 days of flooding, i.e. within the ontogeny of individual plants. The specific recruitment requirements of J. atratus may be a main cause of its rarity in modern Central Europe. In order to prevent regional extinction of J. atratus, we suggest maintaining or re-establishing natural hydrodynamics in the species’ habitats.     

Rareness starts early for disturbance-dependent grassland plant species

Ecological communities always contain a few common species and an abundance of uncommon species. Given that most plant mortality occurs in seeds and seedlings, recruitment success often predicts plant community assemblage and patterning, but observational patterns do not reveal whether plant populations are seed or habitat limited. Grassland plant species make up a sizable portion of the overall native flora in northeastern North America (N.A.), but approximately 30 % of the area’s threatened and endangered flora are grassland species, possibly leftovers from the post-glacial landscape. Yet, close relatives of many rare grassland species thrive in the same range. We investigated whether seed or habitat limitation explained rarity and commonness in remnant grassland species. We used seed addition experiments coupled with microhabitat manipulations (burning and herbivore exclusion) in three different habitat types to evaluate recruitment (germination and seedling survival) limitation for three rare and three common grassland species. Rare grassland species successfully recruited where burning reduced initial competitor density, but seedling survival suggested they were severely limited by interspecific competition. Both the rare and common plant species survived equally well in forest habitats where herbaceous density was low whereas neither survived in the edge habitats. Only the common plants thrived in the high-competition meadow habitat, further suggesting that the rare grassland species are poor competitors. Commonness and rarity are temporal designations that can change as disturbance alters the landscape. Glacial retreat and low precipitation in northeastern N.A. created a landscape suitable for poor competitors that tolerated poor conditions. Our results suggest that rare remnant grassland plants, unlike their close relatives, display more biotic than abiotic limitation as they do not compete well with other plants. These results suggest that suitable habitat is not a spatial location but a temporally transient assemblage of species requirements.     

Propagule pressure and environmental conditions interact to determine establishment success of an invasive plant species,glossy buckthorn (<Emphasis Type="Italic">Frangula alnus</Emphasis>), across five different wetland habitat types

Many invasive plant species are able to establish within a wide range of community types. This establishment success depends on high propagule pressure and successful recruitment of seedlings following propagule dispersal into receptive environments. This study aimed to investigate interactions between propagule pressure and environmental resistance to seedling recruitment of the invasive shrub, glossy buckthorn (Frangula alnus Mill.), over a range of wetland habitat types. We measured propagule deposition using seed traps and recruitment success using sown plots, while characterizing vegetation and abiotic environmental conditions in five adjacent wetland habitat types. Drier habitats, which included Cedar Swamp, Shrub Carr, and String, had lower resistance to buckthorn establishment than the wetter Flark and Cattail Marsh. The drier habitats supported more woody species and provided more raised hummock surfaces essential for successful buckthorn recruitment and establishment. Propagule pressure was also higher in dry habitats that supported higher densities of adult glossy buckthorn, while long-distance dispersal into areas with low adult density was uncommon. Natural recruitment was highest in sites with intense propagule pressure, but experimental sowing of seeds demonstrated that buckthorn establishes in wet sites with higher resistance if propagule pressure is increased and seeds are deposited on hummocks. This study demonstrates the affinity of glossy buckthorn for drier wetland sites, and provides empirical evidence that environmental resistance can be overcome by higher propagule pressure.     

Diplochory in western chokecherry: you can��t judge a fruit by its mesocarp

Western chokecherry (Prunus virginiana var. demissa, Rosaceae) is dispersed by frugivorous birds and carnivores, but it has large seeds that are potentially attractive to rodents that could act as seed predators and dispersers. Here, we quantify the benefits of primary dispersal by birds and secondary dispersal by scatter-hoarding rodents. In the fall, avian frugivores (mostly American robins, Turdus migratorius, and cedar waxwings, Bombycilla cedrorum) consumed 87% of the fruit crop and dispersed 67% of the fruit crop away from parent plants. Rodents removed 89% of seeds that simulated bird-dispersed seed rain from transects in riparian zones and 58% from transects in upland habitats. Rodents scatter-hoarded 91.6% of the seeds they removed, burying most in small caches (two to eight seeds) 8?C25?mm deep. About 39% of the seeds in spring caches produced seedlings. Inside rodent-proof exclosures, 52.1% of seeds buried to simulate rodent caches produced seedlings, 29.7% of which were still alive after 1?year. In contrast, only 3.8% of seeds placed on the soil surface, simulating dispersal by avian frugivores, produced seedlings. Seed dispersal by frugivorous birds likely contributes to colonization of unoccupied habitat through long-range dispersal and to escape from distance-dependent seed mortality near the parent plant. Despite seed losses, rodents offer short-range seed dispersal and bury seeds in more favorable sites for germination, improving seedling emergence and establishment. The combined mechanisms of seed dispersal significantly enhanced chokecherry seedling recruitment by providing more dispersal-related benefits than either frugivorous bird or scatter-hoarding rodents could provide alone.     

Spatial patterns of dispersal, seed predation and germination during colonization of abandoned grassland by Quercus petraea and Corylus avellana

Key processes during recruitment of Quercus petraea and Corylus avellana were investigated in abandoned calcareous grasslands and adjacent scrub using the following methods: (1) observation of hoarding animals during the main period of ripening of acorns and hazel nuts, (2) exposition of seeds on the soil surface and in 5–6 cm depth to test differences in predation and germination, and (3) mapping of seedlings in the grasslands. European jays (Garrulus glandarius) and mice were the main disperser of seeds. Jays preferred acorns, whereas the rodents were less selective, but probably more important for dispersal of nuts. The maximum dispersal distance was about 10–20 m for mice and was estimated several hundred metres for jays. Mice collected hoards of several seeds in about 2 cm depth in the soil, whereas jays stored single seeds. Seed predation and probably hoarding by mice were highest under scrub and in unmown grassland, while jays preferred mown sites for hoarding. However, hiding of seeds in the soil reduced losses in all sites. Predation of nuts was slightly less intensive than that of acorns. Seeds of Corylus were more sensitive to desiccation than Quercus, but in both species germination was lower for seeds exposed on the soil surface and in drier sites. Quercus and Corylus were the most abundant woody species in the fallow grasslands, probably due to the effective multi-staged dispersal by jays and mice, whereas wind-dispersed and fleshy-fruited species were less common; the latter restricted to margins of adjacent scrubland. The study provides several examples for discordance in suitability of patches for seeds and seedlings due to different habitat requirements of successive developmental stages. This result emphasizes the need for studies in the multiple stages during recruitment of vertebrate-dispersed plants.     

Extreme Reproduction and Survival of a True Cliffhanger: The Endangered Plant Borderea chouardii (Dioscoreaceae)

Cliff sides are extreme habitats, often sheltering a rich and unique flora. One example is the dioecious herb Borderea chouardii (Dioscoreaceae), which is a Tertiary, tropical relict, occurring only on two adjacent vertical cliffs in the world. We studied its reproductive biology, which in some aspects is extreme, especially the unusual double mutualistic role of ants as both pollinators and dispersers. We made a 2-year pollination census and four years of seed-dispersal experiments, recording flower visitors and dispersal rates. Fruit and seed set, self-sowing of seeds, seedling recruitment, and fate of seedlings from seeds sowed by different agents were scored over a period of 17 years. The ants Lasius grandis and L. cinereus were the main pollinators, whereas another ant Pheidole pallidula dispersed seeds. Thus ants functioned as double mutualists. Two thirds of all new seedlings came from self-sown seeds, and 1/3 was dispersed by ants, which gathered the seeds with their oil-rich elaiosome. Gravity played a minor role to dispersal. Both ant dispersal and self-sowing resulted in the same survival rate of seedlings. A double mutualism is a risky reproductive strategy, but B. chouardii buffers that by an unusual long–term demographic stability (some individuals exceed 300 years in lifespan) and its presence in a climatically very stable habitat, inaccessible to large herbivores. Such a combination of traits and habitat properties may explain the persistence of this relict species.     

Effect of seed predation on seed bank size and seedling recruitment of bush lupine (Lupinus arboreus)

Whether seed consumers affect plant establishment is an important unresolved question in plant population biology. Seed consumption is ubiquitous; at issue is whether seedling recruitment is limited by safe-sites or seeds. If most seeds inhabit sites unsuitable for germination, post-dispersal seed consumption primarily removes seeds that would otherwise never contribute to the population and granivory has minimal impacts on plant abundance. Alternatively, if most seeds ultimately germinate before they lose viability, there is greater potential for seed consumption to affect plant recruitment. Of the many studies on seed consumption, few ask how seed loss affects seedling recruitment for species with long-lived seed banks. We examined post-dispersal seed predation and seedling emergence in bush lupine (Lupinus arboreus), a woody leguminous shrub of coastal grasslands and dunes in California. We followed the fate of seeds in paired experimental seed plots that were either protected or exposed to rodent granivores in grassland and dune habitats. Significantly more seeds were removed by rodents in dunes than grasslands. In dunes, where rodent granivory was greatest (65% and 86% of seeds removed from plots by rodents in two successive years), there is a sparse seed bank (6.6 seeds m⁻²), and granivory significantly reduced seedling emergence (in the same two years, 18% and 19.4% fewer seedlings emerged from exposed versus protected plots), suggesting seed rather than safe-site limited seedling recruitment. In contrast, rodents removed an average of 6% and 56% of seeds from grassland plots during the same two years, and the grassland seed bank is 43-fold that of the dunes (288 seeds m⁻²). Even high seed consumption in the second year of the study only marginally influenced recruitment because seeds that escaped predation remained dormant. Burial of seeds in both habitats significantly reduced the percentage of seeds removed by rodents. Results suggest that granivores exert strong but habitat-dependent effects on lupine seed survival and seedling emergence. Received: 24 October 1996 / Accepted: 4 February 1997     

Dispersal and microsite limitation of a rare alpine plant

Knowledge on the limitation of plant species’ distributions is important for preserving alpine biodiversity, particularly when the loss of alpine habitats because of global warming or land use changes is faster than colonization of new habitats. We investigated the potential of the rare alpine plant Campanula thyrsoides L. to colonize grassland sites of different suitability on a small mountain plateau in the Swiss Alps. A total of 15 experimental sites were selected according to their differences in habitat suitability for adult C. thyrsoides, which was measured by the Beals index. At each site we applied a disturbance treatment, added seeds at different densities and monitored the survival of seedlings over two consecutive years. The number of surviving seedlings was not positively related to habitat suitability for adult C. thyrsoides. Furthermore, C. thyrsoides appears to be strongly dispersal limited at the regional scale because seed addition to unoccupied habitats resulted in successful germination and survival of seedlings. Since an increase of seed density in already occupied sites did not affect the number of seedlings, we suggest that C. thyrsoides is microsite limited at the local scale. Microsite limitation is supported by the result that seedling survival of the species was enhanced in vegetation gaps created by disturbance. We conclude that C. thyrsoides may become endangered in the future if environmental changes cause local extinction of populations. An appropriate management, such as a disturbance regime for enhancing recruitment in existing populations, may ensure the long-term survival of this rare alpine plant species.     

Rhinanthus minor population genetic structure and subspecies: Potential seed sources of a keystone species in grassland restoration projects

In the last few decades unimproved semi-natural grasslands have been affected by intensification of land use and habitat fragmentation. Because of their biodiversity these species-rich grasslands are of high conservation importance and efforts are under way to restore such habitats. Detailed knowledge of within species diversity will aid deciding on the optimal seed source for such restoration projects, e.g. local genotypes or ecotypes. Rhinanthus minor is a species that is typically found in semi-natural grasslands and is commonly used in grassland restoration projects. This is because R. minor is a hemiparasitic plant that takes minerals and nutrients from its host, which in turn decreases the host's biomass and leads to opportunities for less competitive species in the vegetation. Here, we investigate genetic diversity within and between R. minor populations. This allowed us to test whether the six different subspecies of R. minor that have been described in the UK, based on their morphology, flowering time, and habitat, can be differentiated using molecular markers. We identified moderate levels of genetic differentiation between R. minor populations within the UK. In addition, R. minor individuals from the UK appear to be distinct from R. minor and Rhinanthus angustifolius individuals from other European countries based on microsatellite genotyping and DNA sequencing of cpDNA and rDNA ITS. The molecular markers used in the current study did not separate populations of R. minor based on either their subspecies or habitat. The implication for the use of R. minor in grassland restoration projects seems to be that it is not necessary to use local seeds or seeds from the same subspecies.     

Vertebrate frugivory and seed dispersal of a Chihuahuan Desert cactus

Vertebrate frugivory of fleshy-fruited plants may be very important for the recruitment of sexually derived seedlings if it represents the main mechanism of primary seed-dispersal.Opuntia rastrera produces fleshy fruits rich in water and sugars that are attractive to vertebrate frugivores. However, there is a very low rate of seedling recruitment in natural conditions. One of the causes that can influence this low recruitment is an insufficient seed dispersal due to a low fruit removal even under different resource (fruits) availability. To test this prediction, we studied the production and consumption of fruits in two consecutive years in two vegetation types: nopaleras (dense Opuntia-dominated scrublands) and grasslands with sparse populations of O. rastrera. Plant cover, fruit production and removal, and frugivore identity were recorded within each vegetation type in four randomly selected circular plots (7854 m²). Fruit production per area was higher in nopaleras in both years whereas per cladode production did not differ between vegetation types but differed between years in response to variation in precipitation. Fruit consumption by vertebrates was high (100%) and independent of spatial and temporal fruit availability. The intensity of fruit removal was inversely related to resource availability: it was faster in the less dense community (grassland) and in the driest year. Contrary to other studies with similar cacti, fruit removal by small mammals was insignificant whereas main consumers were birds and large mammals. Vertebrate frugivory represents the only mechanism of primary dispersal of seeds as all fruits are removed in about one month. Despite the high quantity of viable seeds (more than one million per ha in nopaleras and a tenth of that in grasslands) that are dispersed by frugivores after the consumption of about 300 kg of fruit per ha in nopaleras and a tenth of that in grasslands, the rare establishment of seedlings (about one seedling per three million of seeds produced) reported in the literature indicates that the interaction between O. rastrera and the disperser guild is indeed very asymmetrical. We speculate that the harsh conditions for cactus establishment found in this ecosystem demand a high investment in disperser rewards (fleshy fruits) to allow a very modest rate of sexually-derived seedling establishment.     

Paddock‐scale patterns of seed production and dispersal in the invasive shrub Acacia nilotica (Mimosaceae) in northern Australian rangelands

Paddock‐scale Acacia nilotica L. Willd. ex Del. (prickly acacia) spatial distribution, seed production and dispersal patterns were investigated in the Astrebla (Mitchell) grasslands of northern Australia as a step toward predicting future patterns of invasion. A number of hypotheses were tested based on what we know of this species in both Australia and regions where it is native, for example South Africa. It was hypothesized that most A. nilotica seeds would be produced by trees in riparian habitats with access to permanent water. In addition, we predicted that seeds would be dispersed throughout the Astrebla grassland landscape by cattle, following observations that cattle readily ingest and pass seeds and that cattle have access to all areas within paddock boundary fences. Tree density, seed production and seed dispersal by cattle were measured along a series of transects from permanent watering points to paddock boundary fences. Trees associated with permanent watering points produced more seeds per unit area and occurred at higher density than their non‐riparian counterparts. The importance of riparian trees decreased in years with high rainfall and in paddocks with only small areas of riparian habitat. Cattle spread dung and seeds throughout paddocks, with peaks of deposition adjacent to permanent watering points. These results suggest that invasion patterns are likely to be uneven across the landscape and may be reactive to climate. High seedling recruitment and possible thicket formation is expected adjacent to permanent watering points and wherever cattle congregate. Patterns of recruitment in non‐riparian areas are likely to be relatively sparse. The importance of post‐dispersal factors in determining recruitment patterns is discussed.     

Recruitment of two Opuntia species invading abandoned olive groves

In Europe, many agricultural areas are now abandoned and hence can be invaded by exotic species. The abundance and spatial distribution patterns of two Opuntia species were studied in old olive groves in the Parc Natural del Cap de Creus, Catalonia (Spain). Seedling recruitment (97.3% and 51.5% of juveniles for O. maxima and O. stricta, respectively) was higher than recruitment by cladodes. O. maxima had more seedlings recruited beneath olive trees and beneath Opuntia adults than expected. Most O. stricta seedlings were also located beneath Opuntia adult plants. However, although most seedlings were recruited beneath Opuntia, some (10–30%) were found away from putative parental plants. This may be due to seed dispersal by birds and wild boars. Seeds dispersed by wild boars were not significantly more viable than seeds from intact fruits. Seedlings grow very slowly but have a high survival rate. In conclusion, Opuntia seedling recruitment is very successful and ensures the persistence of these species within old olive groves. Consequently, it prevents restoration from an agricultural land-use back to the native community.     

Disturbance,microclimate and historical habitat connectivity determine the population performance of the threatened grassland specialist Carex caryophyllea in remnant grasslands

Many grassland specialist plant populations in Europe have become restricted to remnant habitats. The performance of these populations depends on both species‐specific traits and local and landscape level aspects of habitat quality. Understanding which specific local or landscape level conditions determine the performance of grassland species populations in remnant habitats would help design the restoration of the habitats and to detect the conditions that favour the long‐term persistence of grassland species in them. Such information is especially needed in urbanised landscapes, where remnant habitats engulfed by urban land use types may experience increased erosion, higher temperatures and invasion by alien species. This study investigates the population performance determinants of Carex caryophyllea (VU), a grassland specialist, in 43 remnant grasslands in an urban‐rural gradient in Finland. The population performance was assessed with metrics of persistence, establishment and reproduction, and related to environmental conditions with generalized additive models and redundancy analysis. The most important positive determinants for the performance of C. caryophyllea populations were disturbance through management or ground erosion, a warm microclimate, large habitat area and high historical connectivity to suitable grassland habitats. Present connectivity to other C. caryophyllea populations had a weak and near‐significant positive relationship with population performance. Urbanisation of the surrounding landscape correlated with population performance as well, possibly due to the high historical cover of grasslands in presently urbanised landscapes. The results imply that the most effective restoration method of remnant C. caryophyllea populations would be reinstating disturbance regimes in overgrown habitats with warm microclimates close to suitable habitats and other existing populations, whether urban or rural. This would counteract the species future decline due to possible extinction debts and help the species persist in the study area in the long term.     

Marginal holly (Ilex aquifolium L.) populations in Mediterranean central Spain are constrained by a low-seedling recruitment

Seedling density and survival was studied in holly woodland located in the mountains of the Sistema Central, Madrid (Spain) in 1998. Ten permanent quadrats of 50×50 cm were randomly placed in each of the following habitats: under closed holly canopy, in canopy gaps, at the edge of the holly woodland, in the adjacent open grassland, and under fleshy- and dry-fruited shrubs located in this grassland. Emerged seedlings were periodically monitored and their survival was checked during the following 2 years. Additionally, a set of 20 seedlings were planted in each of the cited habitats and their survival was followed in parallel with the naturally emerged seedlings. The influence of light on holly survival was tested by experimental planting of holly seedlings under six different controlled light intensities.Seedling emergence was strongly habitat dependent, with the highest seedling densities in the closed holly woodland and in canopy gaps, and a complete absence in the grassland. Survival to the first year was nil. No naturally emerged seedling reached 1 year of age, independent of the habitat studied. Only one of the planted seedlings in the field survived to the first year. Despite this, mean seedling lifespan was different between habitats, with canopy gaps, woodland edge, and under fleshy-fruited shrubs being the habitats with highest average lifespan. Seedling response to the distinct light intensities is largely similar to that seen in the different field habitats, but other factors in addition to light, mainly summer drought and disturbances are proposed to explain the 100% mortality observed.Seedling incorporation is the recruitment dynamic bottleneck of this holly population, being more severe here than in other places studied further north. This may explain the increasing rarity of the species in the southern half of the Iberian Peninsula, where Ilex is close to its southern limit of its distribution.     

Forest edges and fire ants alter the seed shadow of an ant-dispersed plant

Exotic species invade fragmented, edge-rich habitats readily, yet the distinct impacts of habitat edges and invaders on native biota are rarely distinguished. Both appear detrimental to ant-dispersed plants such as bloodroot, Sanguinaria canadensis. Working in northeastern Georgia (USA), an area characterized by a rich ant-dispersed flora, fragmented forests, and invasions by the red imported fire ant, Solenopsis invicta , I monitored the interactions between ants and S. canadensis seeds in uninvaded forest interiors, uninvaded forest edges, invaded forest interiors, and invaded forest edges. I observed 95% of the seed dispersal events that occurred within the 60-min observation intervals. Seed collection rates were similar among all four (habitat × invasion) groups. The presence of invasive ants had a strong effect on seed dispersal distance: S. invicta collected most seeds in invaded sites, but was a poorer disperser than four of five native ant taxa. Habitat type (interior versus edge) had no effect on seed dispersal distance, but it had a strong effect on seed dispersal direction. Dispersal towards the edge was disproportionately rare in uninvaded forest edges, and ants in those habitats moved the average dispersed seed approximately 70 cm away from that edge. Dispersal direction was also skewed away from the edge in uninvaded forest interiors and invaded forest edges, albeit non-significantly. This biased dispersal may help explain the rarity of myrmecochorous plants in younger forests and edges, and their poor ability to disperse between fragments. This is the first demonstration that forest edges and S. invicta invasion influence seed dispersal destination and distance, respectively. These forces act independently.     

Why some species cannot colonise restored habitats? The effects of seed and microsite availability

Restoration of calcareous grasslands was promoted as a conservation strategy to reduce the risks imposed by habitat loss and fragmentation. Restoration already provided promising results for several taxa, however some specialist species still fail at colonising restored habitats. We aimed at explaining this lack of colonisation success for three calcareous grasslands specialist species in southern Belgium: Pulsatilla vulgaris; Trifolium montanum; and, Veronica prostrata. We studied: (i) germination in control and outdoor conditions (cold, heat, smoke and litter effects); (ii) in situ seedling emergence patterns (effects of seed addition and germination microsites availability). The three species were able to germinate in Petri dishes in the absence of treatment. Cold enhanced the germination of V. prostrata. Fire-related treatments (heat shock and smoke exposure) did not enhance germination and were deleterious to V. prostrata. Litter cover improved P. vulgaris emergence in outdoor containers, but had a negative effect on V. prostrata. In the field, V. prostrata did not emerge. T. montanum seedlings were observed in the reference grasslands when seeds were added, but not in the restored grasslands. P. vulgaris emerged in the reference grasslands, and to a lower degree in the restored grasslands. The combination of seed addition and microsites availability for seed germination resulted in enhanced seedling emergence for P. vulgaris. Our results suggest that seed and microsite availability can be limiting factors for site colonisation, but the combination of both is likely much more limiting. Lower seedling emergence in restored than in reference grasslands suggests a lower habitat quality in restored grasslands.