In southern humid Chaco forests dominated by Schinopsis balansae, woody plants are clumped and species composition varies spatially over short distances. We examined how these spatial patterns are associated with local environmental heterogeneity for three size classes of woody individuals: adults, saplings, and seedlings. Our study was based on the detailed description of two forest plots (3200 m² in total) in which we mapped all individuals of woody species, and delimited patches with different microrelief, soil moisture, and ground cover of terrestrial bromeliads. Our results showed that woody-individual distribution is related to local environmental heterogeneity. For all size classes, density of woody individuals was highest on convex patches. These convex patches were dominated by both tree (Acacia praecox and Achatocarpus praecox) and shrub species (Celtis pallida, Capparis retusa and Grabowskia duplicata), while level patches were dominated only by tree species (Schinopsis balansae, Prosopis spp. and Geoffroea decorticans). Drier patches contained all the woody species present in the forest, while wetter patches contained only a subset of them. Within convex patches, trees are likely to occur in places with bromeliad colonies and shrubs in patches without bromeliads. The results suggest forest structure is controlled by environmental heterogeneity associated with microrelief and soil moisture.

Nomenclature: Burkart (1969–1987); Digilio (1971–1974); Pérez Moreau (1994).

Abbreviation: MRPP = Multiple Response Permutation Procedure.

We studied canopy structure, shoot architecture and light harvesting efficiencies of the species (photon flux captured per unit above-ground plant mass) in a series of exclosures of different age (up to 4.5 yr) in originally heavily grazed grassland in N Japan. Vegetation height and Leaf Area Index (LAI) increased in the series and Zoysia japonica, the dominant in the beginning, was replaced by the much taller Miscanthus sinensis. We showed how this displacement in dominance can be explained by inherent constraints on the above-ground architecture of these two species. In all stands light capture of plants increased with their above-ground biomass but taller species were not necessarily more efficient in light harvesting. Some subordinate species grew disproportionally large leaf areas and persisted in the shady undergrowth. Some other species first grew taller and managed to stay in the better-lit parts of the canopy, but ultimately failed to match the height growth of their neighbours in this early successional series. Their light harvesting efficiencies declined and this probably led to their exclusion. By contrast, species that maintained their position high in the canopy managed to persist in the vegetation despite their relatively low light harvesting efficiencies. In the tallest stands ‘later successional’ species had higher light harvesting efficiencies for the same plant height than ‘early successional’ species which was mostly the result of the greater area to mass ratio (specific leaf area, SLA) of their leaves.

This shows how plant stature, plasticity in above-ground biomass partitioning, and architectural constraints determine the ability of plants to efficiently capture light, which helps to explain species replacement in this early successional series.

Nomenclature: Makino (1962); Ohwi (1965).

Abbreviations: LAI = Leaf area index; LAR = Leaf Area Ratio; LMR = Leaf Mass Ratio; PPFD = Photosynthetically active photon flux density; SLA = Specific Leaf Area.

In late successions of the boreal forest of northern Sweden the evergreen dwarf shrub Empetrum hermaphroditum forms an extensive cover and is believed to spread mainly vegetatively through layering. To analyse the process of population establishment and the relative importance of sexual vs asexual reproduction and the spatial clonal distribution of this species we selected one mainland and two island sites of different post-fire successional ages (145, 375 and 1720 yr since last fire, respectively). Using 61 polymorphic RAPD markers, we found 96 genotypes in a total of 133 samples. All three populations showed high levels of genetic variation. AMOVA analysis revealed that 33% of the total variation resided among sites, 26% resided among plots within sites and 41% was due to variation within plots. The youngest population had only 14% clonal fraction. In contrast, the oldest population had > 30% clonal fraction and many genets had dimensions of 10–40 m and were intermingled. It appears that E. hermaphroditum establishes by seeds to a larger extent than previously thought and that the clonal spread by layering is rather slow.

Nomenclature: Haapasaari (1988).

Abbreviation: PCR = Polymerase Chain Reaction; RAPD = Random Amplified Polymorphic DNA.

The artificial seedling assay method has been used in several tropical and temperate forests to estimate seedling damage and subsequent mortality due to non-trophic microdisturbances such as litterfall and trampling. However, there has been no evidence presented to support the assumption that artificial seedling damage correlates with natural seedling mortality. In this study we evaluated the use of artificial seedlings in five New Zealand forests by comparing damage rates of artificial seedlings with damage and mortality rates of an equal number of selected natural seedlings. A total of 1200 artificial, and 1200 natural seedlings were monitored monthly for two years.

Litterfall damage rates of natural seedlings were correlated with those of artificial seedlings. However, there was no relationship found between artificial and natural seedling damage due to animals, or between artificial seedling damage and natural seedling mortality for either cause. Artificial seedlings should not therefore be used to estimate seedling mortality due to non-trophic microdisturbance. Monitoring natural seedlings enabled non-trophic animal damage to be readily detected and provided additional information on animal-plant interactions.

The value of using artificial seedlings is that they provide a measure of the litterfall disturbance potential that is independent of the patchy distribution of natural seedlings within safe sites, and independent of species specific resistance to damage. They therefore provide an indication of the selective pressure on seedlings due to litterfall and are useful for comparing the relative ‘safety’ of different forests and microsites. Monitoring natural seedlings provides a measure of the damage and mortality due to litterfall for a specific seedling size class. Concurrently monitoring artificial and natural seedlings for litterfall disturbance provides more information than the same effort expended on either method alone.

Nomenclature: Allan (1961) amended by Connor & Edgar (1987).

Soil resource availability may affect plant regeneration by resprouting in disturbed environments directly, by affecting plant growth rates, or indirectly by determining allocation to storage in the resprouting organs. Allocation to storage may be higher in stressful, low resource-supply soils, but under such conditions plant growth rates may be lower. These factors could act in opposite directions leading to poorly known effects on resprouting. This paper analyses the role played by soil resources in the production and growth of resprouts after removal of above-ground plant tissues in the Mediterranean shrub Erica australis. At 13 sites, differing in substrate, we cut the base of the stems of six plants of E. australis and allowed them to resprout and grow for two years. Soils were chemically analysed and plant water potential measured during the summer at all sites to characterize soil resource availability. We used stepwise regression analysis to determine the relationships between the resprouting response [mean site values of the number of resprouts (RN), maximum length (RML) and biomass (RB)] and soil nutrient content and plant water potential at each site. During the first two years of resprouting there were statistically significant differences among sites in the variables characterizing the resprouting response. RML was always different among sites and had little relationship with lignotuber area. RN was less different among sites and was always positively correlated with lignotuber area. RB was different among sites after the two years of growth. During the first months of resprouting, RN and RML were highly and positively related to the water status of the plant during summer. At later dates soil fertility variables came into play, explaining significant amounts of variance of the resprouting variables. Soil extractable cations content was the main variable accounting for RML and RB. Our results indicate that resprout growth of E. australis is positively affected by high water availability at the beginning of the resprouting response and negatively so by high soil extractable cation content at later periods. Some of these factors had previously shown to be related, with an opposite sign, to the development of a relatively larger lignotuber. Indeed, RML and RB measured in the second year of resprouting were significantly and negatively correlated with some indices of biomass allocation to the lignotuber at each site. This indicates that sites favouring allocation to the resprouting organ may not favour resprout growth.

Abbreviations: AB = Above-ground biomass; Cat_extr = soil extractable cations; FB = Foliar biomass; LA = Lignotuber area; LB = Lignotuber biomass; N_tot = Soil total nitrogen; P_avail = Soil available phosphorus; RA = Root basal area; RB = Resprout biomass; RML = Resprout maximum length; RN = Resprout number; Ψ_pd = Predawn plant water potential.

Nomenclature: Castroviejo et al. (1986).

Large succulent leaf rosettes are a characteristic life form in many deserts. In certain areas they become the dominant life form, creating a vegetation type indicated as rosette scrub. The large number of rosette species suggests a close relationship between form and environment. Rosettes are excellent harvesters of low-intensity rains and fogs. We propose that some rosette-dominated formations of the Mexican mountains, namely the montane rosette scrub, occur in altitudinal belts around mountains where fog is abundant. We sampled four altitudinal gradients in mountains with different flora recording the abundance and richness of plants. At one site, the Tehuacán Valley, we also measured the quantity of fog along the gradient, below, above and in the rosette scrub for one year. We found that the abundance and richness of succulent rosette species are strongly associated with altitude, showing maximum values in the well-defined elevational belts where the montane rosette scrub occurs. Other life forms, such as stem succulent cacti or woody shrubs, do not show this mid-elevation pattern. The altitudinal ranges where the montane rosette scrub occurs usually coincide with areas where clouds and fog form. Our micrometeorological measurements indicate that rosette plants growing within a cloud belt can increase their water supply by 10–100% by harvesting fog. Outside these belts fog harvest is negligible. Desert rosettes constitute one of the most common fog-harvesting growth forms.

Nomenclature: Gentry (1982) for Agavaceae, Smith & Downs (1974) and Roberts (1989) for Bromeliaceae.

In heavily altered landscapes, where vegetation is not natural and where people are the main source of ignitions, relationships between fire occurrence and climate conditions may be unclear. The objective of this study was to evaluate to what extent territories with similar Potential Natural Vegetation (PNV) in peninsular Spain differ in their forest fire characteristics. From 1974 to 1994, more than 174000 fires occurred. We used (1) the Spanish data base of forest fires, (2) a PNV map and (3) a land use map. Separate fire characteristics, based either on the number of fires occurred or the area burned, were obtained for each of the ca. 5000 grid-cells (10 km × 10 km) into which peninsular Spain is divided in the UTM projection. Also, meteorological conditions at the time of fire ignition, cause of ignition and present forest cover were referred to the same grid-cells as external factors potentially determinant of fire occurrence. The relationships between fire regime characteristics and PNV units were explored with Principal Components Analysis (PCA). The role of the three sets of external factors in the fire characteristics was evaluated with Redundancy Analysis (RDA). Groups of similar PNV types were clearly segregated, suggesting a gradient of fire characteristics. Higher fire incidence (higher frequencies and spatial incidence of fires, but lower proportions of grid-cells affected by large fires) was associated with Atlantic, warm territories with deciduous forests as PNV. Intermediate fire frequency and rotation period, but with a higher relative incidence of medium and large fires occurred in Mediterranean PNV units, dominated by sclerophyllous oak forests. Low fire frequency and long rotation periods, with strong seasonal and yearly variability occurred for PNV units in the cold uplands (Fagus, Pinus, Abies, Juniperus) or in the semi-arid, shrubby PNV units. The cause of ignition best explained the patterns of forest fire characteristics, followed by weather conditions.

Our results indicate that, even in human influenced regions, climate and soil conditions exert control on the resulting forest fire characteristics, as indicated by the high segregation of the PNV types. However, the role of man was crucial in shifting the patterns of fire incidence. This was so that highest fire incidence occurred in regions that, otherwise, would be expected to have a much lower one, thus posing a serious threat for such areas. PNV maps, by providing a phytogeographical framework for characterizing forest fires, could be valuable tools for applying research results to forest fire management policies, taking properly into account the underlying determinant factors.

Abbreviations: PCA = Principal Components Analysis; PNV = Potential Natural Vegetation; RDA = Redundancy Analysis.

The clonal sedge Carex humilis forms rings of densely aggregated ramets in a dry grassland community in Central Europe. We describe the small-scale spatial variation, both in abiotic factors and vegetation, in relation to these rings. Compared to the surrounding vegetation the cover of plants, other than C. humilis, was significantly lower both in the central area of rings and within the rings themselves. The vegetation structure was also different. The soil was more fertile in the central area and within the ring than in the surroundings, measured both directly and by the abiotic response values of the vascular plants. We conclude that neither resource depletion nor competition from other plants were likely to be responsible for the low ramet density in the central area of C. humilis rings. Instead, we suggest that the ring form is caused either by the deposition of growth inhibiting substances or by intrinsic morphological rules.

Nomenclature: Tutin et al. 1964–1993.

We characterized the abundance, size and spatial patterning of canopy gaps, as well as gap-forming processes and light availability in boreal, sub-boreal, northern temperate and subalpine old-growth forests of northwestern British Columbia. The proportion of area in canopy gaps ranged from 32% in northern temperate forests to 73% in subalpine forests. Evenly distributed developmental gaps were dominant but permanent openings created by edaphic components and by shrub communities were also common, particularly in sub-boreal forests. Abundant gaps, large gap sizes, high numbers of gap makers per gap and frequent gap expansion events suggest that gaps have long tenure in these forests. Snapped stems and standing dead mortality were the most common modes of mortality in all forest types resulting in little forest floor disturbance, creating few germination sites for seedling establishment. We found high mean light levels (16–27% full sun) and little difference between non-gap and gap light environments. Our results suggest that gap dynamics in these forests differ fundamentally from those in temperate and tropical forest ecosystems.

Nomenclature: Hitchcock & Cronquist (1994).

In this paper we analyse six communities with seven Mimosa species in the Tehuacán-Cuicatlán valley, Mexico. All species are endemic to Mexico and four are endemic to the valley. Mimosa species are found in (1) the ‘matorral xerófilo’ (arid tropical scrub): Mimosa calcicola, M. lacerata, M. luisana, M. polyantha and M. purpusii, and (2) the ‘selva baja caducifolia’ (tropical deciduous forest): M. adenantheroides, Mimosa texana var. filipes. Most of them occur in similar soil environments, while M. polyantha and M. calcicola establish in particular soil conditions and only M. luisana establishes in two different sites showing a wider range of adaptation to soil characteristics. The communities studied include 24 plant families, 51 genera and ca. 70 species (5% of the total flora estimated in the valley). Heterogeneity was found among the communities. Our results point to the replacement of ‘matorral xerófilo’ and ‘selva baja caducifolia’ by ‘matorral espinoso’ (thorny scrub). Thorny species (e.g. Acacia cochliacantha, Mimosa spp.) are becoming the dominant/codominant elements in the communities. Within the communities, Mimosa species have a significant influence on soil pH, organic matter and electrical conductivity values. These species contribute to ameliorate soil nutrient conditions (OM, N_tot, P, Ca, Mg, Na and K contents) as well as other environmental factors (e.g. temperature, shade) under their canopy in degraded plant communities and may serve as ‘resource islands’.

Nomenclature: Dávila et al. (1993); for endemic Mimosa species, see ^{Material and Methods}.

A cross-over fertilization experiment was carried out in Dutch floating fens to investigate effects on biomass production in the same and the following years. In total 16 fertilizer treatments were applied, combining four treatments in 1999 with four treatments in 2000 (addition of 20 g.m⁻² N, 5 g.m⁻² P, both elements and unfertilized control). The above-ground biomass production of vascular plants was co-limited by N and P in both years. However, in plots that were only fertilized in 1999 the effects of individual nutrients differed between the two years: N-fertilization slightly increased the amount of biomass produced in the same year (1999), whereas P-fertilization did so in the following year (2000). Fertilizer applied in 1999 also influenced the effects of fertilizer applied in 2000. One year after N-fertilization vascular plant growth was still co-limited by N and P, but one year after P-fertilization, vascular plant growth was only limited by N. Bryophyte biomass responded weakly to fertilization. Nutrient concentrations in plant biomass, nutrient standing crops and measurements of N and P availability in the soil indicated that one year after fertilization, the N-fertilizer had mostly ‘disappeared’ from N-fertilized plots, whereas the availability of P remained markedly enhanced in P-fertilized plots. In addition, P-fertilization enhanced the uptake of N by plants the following year. The time-dependence of fertilizer effects was probably caused by (1) higher addition of P than of N relative to the requirements of plants; (2) longer retention of P than of N in the system; (3) positive effect of P-fertilization on the availability of N; (4) contrasting effects of N- and P-fertilization on nutrient losses by plants and/or on their responses to subsequent nutrient addition; (5) changing interactions between vascular plants and mosses (mainly Sphagnum spp.); (6) nutrient export through the repeated harvest of above-ground biomass. To determine which nutrient limits plant growth fertilization experiments should be short, avoiding that indirect effects of a non-limiting nutrient influence results. To indicate how changed nutrient supply will affect an ecosystem longer-term experiments are needed, so that indirect effects have time to develop and be detected.

Nomenclature: Tutin et al. (1964–1993).

We investigated the effects of dispersal limitation, diaspore density, soil chemistry, disturbance and mollusc herbivory on the abundance and distribution of the perennial forest herb Dentaria bulbifera (Brassicaceae). After experimental sowing of bulbils in originally empty patches, juveniles of Dentaria emerged in 83% of plots. The seventh year after sowing, plants still persisted and were reproducing in several plots, demonstrating that the species is dispersal-limited. However, long-term survival of transplants was higher at sites where Dentaria occurred naturally, and this difference increased with time after sowing. Thus, the importance of local factors may only be expressed fully several years after colonisation. Differences in soil chemistry and disturbance treatment were not associated with differences in patch suitability. Removal of molluscs, however, significantly increased recruitment, and the effects of a single molluscicide treatment persisted after four years. We conclude that dispersal is not the only limitation to the distribution of Dentaria, and that local factors, such as mollusc herbivory, are also crucial.

Nomenclature: Mossberg et al. (1992).

In southern Mediterranean Pinus sylvestris forests there are grassy meadows that resist invasion of trees despite the proximity to seed sources. In this study, we investigate the mechanisms blocking colonization by Pinus sylvestris of the meadows. Two experiments were conducted in which seeds were sown either at 1 cm depth or on the surface to simulate dispersal, and three treatments of vegetation removal were applied: Disturbed (where the herbaceous layer was eliminated, exposing the mineral soil), Clipped (vegetation cut at ground level) and Control (no disturbance of the herbaceous layer). In addition, the effect of seed predators was controlled by using wire cages in the case of the surface sown experiment.

When seeds were sown at 1 cm depth, seedling emergence was not reduced by the herbaceous layer. In contrast, when seeds were surface sown and predators were excluded, the rate of emergence was low in the Control treatment, intermediate in Clipped and high in Disturbed. Seedling emergence was, however, minimal when predators were not excluded, irrespective of the disturbance level. Seedling survival and growth after three years of study were similar among treatments. The results show that the seed predation and the physical barrier created by the herbaceous layer are the two mechanisms blocking the encroachment of Pinus sylvestris onto these Mediterranean mountain meadows, limiting the regeneration and potential expansion of the forest.

Nomenclature: Tutin et al. (1964–1980).

A semi-natural grassland in Sweden was exposed to an elevated CO₂ concentration during a six-year open-top chamber experiment. Vegetation composition was assessed twice a year using the point-intercept method. The field had been grazed previously, but when the experiment started this was replaced with a cutting regime with one cut (down to ground level) each year in early August. From the third to the sixth year of the study the harvested material was divided into legumes, non-leguminous forbs and grasses, dried and weighed.

Elevated CO₂ had an effect on species composition (as analysed by Principal Component Analysis) that increased over time. It also tended to increase diversity (Shannon index) in summer, but reduce it in spring. However, the effects of the weather and/or time on species composition and diversity were much more prominent than CO₂ effects. Since the weather was largely directional over time (from dry to wet), with the exception of the fifth year, it was difficult to distinguish between weather effects and changes caused by a changed management regime.

In all treatments, grasses increased over time in both mass and point-intercept measurements, whereas non-leguminous forbs decreased in mass, but not in point-intercept measurements. Legumes increased in the point-intercept measurements, but not in biomass, at elevated CO₂, but not in the other treatments.

Overall, we found that elevated CO₂ affected species composition; however, it was only one of many factors and a rather weak one.

Nomenclature: Krok & Almquist (2001).

Plants associated with traditional agricultural landscapes in northern Europe and Scandinavia are subjected to drastic habitat fragmentation. In this paper we discuss species response to fragmentation, against a background of vegetation and land-use history. Recent evidence suggests that grassland-forest mosaics have been prevalent long before the onset of human agriculture. We suggest that the creation of infield meadows and outland grazing (during the Iron Age) increased the amount and spatial predictability of grasslands, resulting in plant communities with exceptionally high species densities. Thus, distribution of plant species in the present-day landscape reflects historical land-use. This holds also when traditional management has ceased, due to a slow response by many species to abandonment and fragmentation. The distribution patterns are thus not in equilibrium with the present habitat distribution. Fragmentation influences remaining semi-natural grasslands such that species density is likely to decline as a result of local extinctions and invasion by habitat generalists. However, species that for a long time have been subjected to changing mosaic landscapes may be more resistant to fragmentation than is usually believed. Conservation should focus not only on ‘hot-spots’ with high species richness, but also consider species dynamics in a landscape context.