Question: What are the relative effects of resource availability and heterogeneity on the slope of the species-area curve along a floodplain-upland gradient?

Location: Réserve Léon-Provancher, Québec, Canada.

Methods: The mean and coefficient of variation of several environmental variables (soil organic matter, N and pH; irradiance; elevation) were estimated in 70 plots along a flood-plain-upland gradient. The slope of the species-area curve was calculated from subplots nested within each plot. Path analysis was used to determine the effects of the environmental variables on the slope of the species-area curve.

Results: Spatial variation in flooding intensity created a complex gradient along which resource availability, but not resource heterogeneity, varied. Together, the environmental variables explained 73% of the variance in the slope of the species-area curve. Elevation, soil variables (but not light) and heterogeneity in soil fertility (but not in light) were all significantly associated with the slope of the species-area curve. Resource availability explained three times as much of the variance as resource heterogeneity.

Conclusion: The expected positive relationships between resource availability, resource heterogeneity and diversity were obscured by effects of flooding stress and species dominance. Flooding intensity varies inversely with elevation and restricts the pool of species that can occupy the more fertile/heterogeneous section of the gradient, i.e. near the river. The few species combining flooding stress tolerance and competitive ability use the abundant resources near the river to grow quickly and utilize space, thereby enhancing the negative effect of the flooding stress on diversity.

Nomenclature: Marie-Victorin (1995).

Abbreviations: CV = Coefficient of variation; GFI = Goodness of fit index; GLM = Generalized linear model; OM = Soil Organic matter; PPFD = Photosynthetic photon flux density; RMSEA = Root mean square error of approximation.

Questions: Are ordinal data appropriately treated by multivariate methods in numerical ecology? If not, what are the most common mistakes? Which dissimilarity coefficients, ordination and classification methods are best suited to ordinal data? Should we worry about such problems at all?

Methods: A new classification model family, OrdClAn (Ordinal Cluster Analysis), is suggested for hierarchical and non-hierarchical classifications from ordinal ecological data, e.g. the abundance/dominance scores that are commonly recorded in relevés. During the clustering process, the objects are grouped so as to minimize a measure calculated from the ranks of within-cluster and between-cluster distances or dissimilarities.

Results and Conclusions: Evaluation of the various steps of exploratory data analysis of ordinal ecological data shows that consistency of methodology throughout the study is of primary importance. In an optimal situation, each methodological step is order invariant. This property ensures that the results are independent of changes not affecting ordinal relationships, and guarantees that no illusory precision is introduced into the analysis. However, the multivariate procedures that are most commonly applied in numerical ecology do not satisfy these requirements and are therefore not recommended. For example, it is inappropriate to analyse Braun-Blanquet abudance/dominance data by methods assuming that Euclidean distance is meaningful. The solution of all problems is that the dissimilarity coefficient should be compatible with ordinal variables and the subsequent ordination or clustering method should consider only the rank order of dissimilarities. A range of artificial data sets exemplifying different subtypes of ordinal variables, e.g. indicator values or species scores from relevés, illustrate the advocated approach. Detailed analyses of an actual phytosociological data set demonstrate the classification by OrdClAn of relevés and species and the subsequent tabular rearrangement, in a numerical study remaining within the ordinal domain from the first step to the last.

Abbreviations: AD = Abundance/Dominance; CL = Complete Link; DC = Coefficient of Discordance; ED = Euclidean distance; O = Ordinal; M = Metric; NMDS = Non-metric Multidimensional Scaling; OC = Ordinal Clustering; SL = Single Link; UPGMA = Unweighted Pair Group Method or Group Average Clustering.

Question: We were interested if and how variation in frequency and/or size of disturbances affect the dynamics of a montane old-growth forest in Central Europe.

Location: The forest, co-dominated by Fagus sylvatica, Picea abies and Abies alba, is located in Lower Austria and represents one of the few sizable virgin forests in Central Europe.

Methods: We extracted cores from 100 trees using systematic grid sampling (grid cell size 10 m × 10 m) on each of four 1-ha plots distributed across the old-growth remnant of 300 ha. We inferred disturbance events from rapid early growth and release events. For defining release criteria, we applied the boundary line method. We investigated the spatial structure of current age and gap distributions and past disturbance events in grid cells, using a pair density statistic.

Results: The disturbance histories indicate decades with peaks and also extended periods without disturbance. Some peaks occurred synchronously at three of the four plots (1910s, 1930s, 1960s and 1980s). Peaks and gaps in the disturbance chronologies widely agreed with peaks and gaps in the age distributions. Most disturbance events during single decades showed a random spatial distribution.

Conclusions: There is considerable variation in disturbance frequency and/or severity over time. Most disturbance events will rather thin the stand than clear larger areas at once. Following scattered disturbance two pathways occur: (1) gap expansion leading to creation of larger gaps, and (2) gap closure by lateral encroachment or by subcanopy trees growing into the canopy.

Abbreviation: PSP = Permanent sample plot.

Question: How has the terrain, namely elevation, slope and aspect, controlled the cultural modification of the forest-grassland transition in southern Alberta?

Location: The aspen parkland region of the Rocky Mountains foothills southwest of Calgary, Alberta, Canada.

Methods: Dominion Land Survey data, historical maps and a DEM (digital elevation model) were used to determine which decision rules settlers used to plough native grassland and parkland at the turn of the last century. Comparative measurements between historical and current vegetation patterns and GIS techniques were used to explore relationships between terrain, land use and the modification of the landscape.

Results: Grassland at lower elevations, aspen parkland at mid elevations and closed forest at higher elevations dominated the pre-settlement forest-grassland transition. Elevation, hillslope angle and aspect affected which parts of the landscape were cleared and ploughed for agriculture or left in their natural state. Clearing and ploughing occurred mostly on gentle hill-slopes (≤ 6°). At higher elevations with fewer gentle hillslopes clearing and ploughing were limited. Ca. 90% of the original vegetation at lower elevations has been cleared or ploughed, compared to ca. 30% at higher elevations. Almost all of the grassland at lower elevations has been converted to cropland. In the mid-elevation parkland 48% of the tree cover and > 72% of the native grass have been converted to domesticated grass and cropland. Tree cover has expanded at the expense of grass on steep hillslopes and in ravines that presumably were burned frequently by wildfires in pre-settlement times. At higher elevations, the closed forest has remained more or less intact.

Conclusions: Terrain structure has controlled the arrangement of suitable agricultural land and consequently determined the current vegetation patterns in the forest-grassland transition zone in southern Alberta. Tree invasion into grass has been minimal, and has been restricted to hilly topography or wet areas, even though the parkland has expanded somewhat in its distribution. Any effect the cessation of wildfires may have had on aspen expansion has been overshadowed by agriculture, which has converted vast areas of native grassland and aspen parkland to farmland. Thus, along this forest-grassland transition, the strong elevation, slope and aspect have maintained some of the regional variation in vegetation.

Nomenclature: Moss (1982).

Question: Is Rao's quadratic entropy a suitable measure of functional diversity if several traits are considered?

Methods: It is checked whether Rao's quadratic entropy (FD_Q) satisfies a priori criteria suggested by Mason et al. A real data set is used to show that there are often zeros in abundance distributions which maximize functional diversity.

Results and Conclusion: FD_Q fulfils all a priori criteria and it surpasses other proposed indices, because it includes species abundances and more than one trait. Therefore, it seems to be an improvement compared to measures of functional diversity that are currently available. An unexpected property of FD_Q is that its value may decrease if species richness increases. The reason is that functional diversity is influenced by both species-abundance based diversity and differences among species. Introduction of a new species into the community increases the species-abundance based diversity, while it may decrease the average dissimilarity among species.

Abbreviation: FD_Q = Functional diversity measured by Rao's quadratic entropy.

Questions: The 1990s were the warmest decade since the beginning of climate measurements. Based on almost 100 years of monitoring in the Swiss Alps, we asked (1) whether the extraordinary warm climatic conditions of the 1990s are reflected in the floristic composition of Alpine summit vegetation and, if so, (2) what the magnitude and rate of species change has been over the last few decades compared to the documented increase in species richness within the first 80 years of the 20th century.

Location: Ten high mountain summits of the Bernina area in the southeastern Swiss Alps.

Methods: Resurvey of the floristic composition of the uppermost altitudinal 10 m of these summits, applying the same methodology of former two surveys (1905 and 1985) and recording the presence of all vascular plant species.

Results: Whereas the continued increase in plant species richness of high alpine summit vegetation is confirmed, our results also suggest an acceleration of the trend in the upward shift of alpine plants.

Conclusion: Vegetation change in the southeastern Swiss Alps has accelerated since 1985, consistent with a climate change explanation.

Nomenclature: Lauber & Wagner (1998). Verification of synonyms was based on Hess et al. (1984) and the Swiss Web Flora ( www.wsl.ch/land/products/webflora/floramodul1-en.html).

Question: Does the interplay of life-history strategy and tree competition promote tree species co-existence? Using a growth dynamics model, we investigated mechanisms of co-existence among major tree species in a sub-alpine old-growth coniferous forest.

Location: Sub-alpine old-growth coniferous forest at 1850–1920 m a.s.l. in the Ontake Forest Reserve of central Japan.

Methods: We investigated the growth and mortality rates of trees ≥ 5.0 cm stem DBH and recruitment processes in a 2-ha study plot, and developed a model for individual growth that incorporated both intra and interspecific competition and analysed the direction and degree of competitive effect.

Results: Four species, Picea jezoensis var. hondoensis, Tsuga diversifolia, Abies mariesii and Abies veitchii co-occurred as dominant species in the canopy layer. P. jezoensis var. hondoensis and T. diversifolia had low stem densities and bell shaped DBH distributions. In contrast, A. mariesii and A. veitchii had high stem densities and inverse J-shaped DBH distributions. The growth of the species with inverse J-shaped DBH distributions (Abies spp.) was governed by the abundances of the species with bell shaped DBH distributions (Picea and Tsuga). However, Picea and Tsuga were inferior to Abies spp. in terms of recruitment rate (the number of juveniles that grow up to 5.0 cm DBH per year). Therefore, it was suggested that there was a trade-off between recruitment pattern and interspecific competition for species co-existence.

Conclusions: Picea and Tsuga, with inferior recruitment, coexisted with Abies spp., with superior recruitment, by suppressing the growth of potential successors of Abies spp. The interplay of life-history strategies (recruitment and longevity) and interspecific competition therefore plays an important role in promoting species co-existence in this sub-alpine old-growth coniferous forest.

Abbreviations: GLM = General linear model.

Nomenclature: Ohwi & Kitagawa (1992).

Questions: Is dormancy selected as a means by which a seed defers germination until suitable environmental conditions occur? Are plant species in frost-free and drought-free environments less likely to have some form of dormancy than species from environments with frost and/or drought? Are there any relationships between dormancy and three important plant traits: seed size, plant longevity, and growth form?

Methods: Data from the literature including 3164 species were compiled. We quantified the relationship between dormancy and seed size (1795 species), plant longevity (183 species), and growth form (3164 species).

Results: Plant species of environments with frost and/or drought were more likely to possess some form of seed dormancy (morphological, physiological, morphophysiological, or physical) than species in more benign environments. Seeds exhibiting some form of dormancy were lighter in weight and generally more prevalent among herbaceous species. Woody plants had a larger proportion of non-dormant seeds. There was a broad overlap in seed size between species with dormant and non-dormant seeds. We also found no significant relationship between seed dormancy and plant longevity.

Question: Monocot and dicot perennial herbaceous species can be considered two a priori guilds, which interact in regularly mown grasslands. Complementary resource use in space is important for diversity and productivity of perennial herbaceous communities, yet small-scale spatial patterns of guilds and niche complementarity have rarely been linked explicitly. We ask: 1. Do guild interactions in our community generate small-scale patterns such as guild proportionality? 2. Do these patterns coincide with overyielding of guilds, that is, can small-scale spatial patterns of guilds be experimentally linked to niche complementarity?

Location: Georgia, central Caucasian mountains.

Methods: We sampled randomly placed small quadrats and used presence-absence data to measure small-scale spatial patterns within plots. We experimentally tested overyielding of monocots and dicots grown separately and in mixtures.

Results: We found measurable small-scale pattern of guilds in our community: monocots and dicots showed significant guild proportionality. However, variance deficit in the number of species per small quadrat was stronger in monocots than in dicots. Experiments found considerable overyielding, which was unaffected by water stress.

Conclusions: Coincidence of measurable small-scale arrangements of guilds with robust overyielding indicates that spatial arrangement is an important manifestation of niche complementarity. Stronger variance deficit of monocot species at small scale suggests that monocots compete with each other more strongly and, as a consequence, are more regularly dispersed at small spatial scale. Conversely, morphologically more diverse and more plastic dicots may more easily fit within the neighbourhood patchwork created by monocots.

Question: Two questions were posed: 1. Can an independent measure of relative competitive ability be used to predict the abundance of species in mixtures? 2. Is the success of those predictions affected by low fertility (stress simulation) or clipping (disturbance simulation)?

Location: Greenhouse at Carleton University, Ottawa, Canada.

Methods: We collected adult plant ramets of 11 species from the field and transplanted one ramet of each species into 56 containers of 60 L. We applied a 4 × 2 factorial combination of fertilization (none, full nutrients except N, full nutrients except P, full nutrients) and clipping (no clipping, clipping to 10 cm above soil) with seven replicates of each treatment. After two growing seasons the above- and below-ground biomass of each species was determined.

Results: Regression analyses uncovered a significant positive relationship between plant biomass (measured in this study) and relative competitive ability (as measured in an independent study) under all experimental conditions. Both the mean slope and mean R² were lowest in treatments with low nutrients and highest in the full nutrient treatment (irrespective of clipping).

Conclusions: Our results show that (1) at high fertility, relative competitive ability can generally predict the abundance of species in experimental plant communities, and (2) the intensity of competition (inferred from the magnitude of the slope or R²) increased with increasing nutrient supply, particularly nitrogen.

Nomenclature: Gleason & Cronquist (1991).

Question: Do species from communities with different flooding dynamics differ in seed buoyancy? Is there a trade-off between seed buoyancy and seed longevity?

Methods: Seeds of 55 freshwater wetland species were collected and related to communities along the hydrological gradient, ranging from constantly flooded reed beds, through tall herb fens and rich fens, to rarely flooded wet meadows. Species were experimentally tested for seed buoyancy over 210 days in standing water and moving water. Seed longevity for each species was calculated.

Results: Community occurrence along the hydrologic gradient was mirrored by seed buoyancy. Seed buoyancy was highest for species of almost permanently inundated reed beds and lowest for species of rarely inundated wet meadows. The seed buoyancy characteristics of species of reed beds were best adapted to hydrological dynamic conditions mimicked by the treatment moving water and were the most independent of the flow rate of the water. A high percentage of the species used had a transient seed bank. Together with high buoyancy in many species, this suggests a trade-off between dispersal capacity and seed persistence.

Conclusions: In freshwater wetlands with rather stable water levels, many species lack a persistent seed bank and depend on dispersal for colonization of new habitats. Seed buoyancy enhances aquatic seed dispersal, and can be of great importance in both vegetation dynamics and restoration. In wetlands that are inundated for a long period during the year and with a water level above soil surface, higher seed buoyancy enhances the possibility for hydrochory.

Nomenclature: van der Meijden (1996) for phanerogams; Schaminée et al. (1995, 1996) and Stortelder et al. (1998) for syntaxa.

Questions: 1. What are the spatial patterns of all trees, individual tree species, trees within particular height classes, all gaps and gaps with specific properties across the study site in broad-leaved deciduous forest at a range of scales? 2. Are patterns of the above features spatially associated? 3. Are these patterns indicative of gap creation mechanisms and phases of regeneration?

Location: Frame Wood, New Forest, UK.

Methods: Ripley's K-function analysis was applied to spatial information derived from airborne remotely sensed imagery to characterize the patterns of trees and gaps and to test for spatial interactions between these patterns. The patterns of trees and gaps with specific physical and spatial properties were analysed.

Results: The pattern of all tree species combined was random for most scales; Quercus robur followed the same random pattern, while Fagus sylvatica and Betula pendula were clustered over most spatial scales. Large gaps (> 250 m²) and larger trees (> 17.5 m) were randomly distributed, while smaller gaps and smaller trees were clustered. Significant spatial relationships were demonstrated between the patterns of different tree species and between trees within different size classes, as well as between the patterns of trees and gaps with specific properties.

Conclusions: Small gap patterns and field evidence indicated that progressive gap enlargement is the most likely creation mechanism for large gaps (> 250 m²). Clustered patterns of younger individuals were indicative of patches of past regeneration. As a complement to field-based data, data derived from remotely sensed imagery provides spatially comprehensive information with which to further investigate woodland stand/community processes and gap dynamics.

Abbreviations: ATM = Airborne thematic mapper; CHM = Canopy height model; CSR = Complete spatial randomness; GHD = Gap height diversity; GSCI = Gap shape complexity index; LiDAR = Light detection and ranging.

Question: How have the spatial and temporal aspects of past disturbance shaped the current structure and composition of old-growth Picea rubens forests?

Location: Northern Maine, USA.

Methods: We established three 50 m × 50 m plots in old-growth Picea rubens forests and mapped the location of trees and saplings. We extracted increment cores from canopy trees, and recorded growth releases indicating past disturbance. By linking spatial data (tree positions) and temporal data (dated growth releases), we reconstructed the location and size of former canopy gaps back to 1920, and determined a more general disturbance chronology extending as far back as 1740.

Results: We found no evidence of stand-replacing disturbances. The disturbance dynamic includes pulses of moderate-severity disturbances caused by wind storms and host-specific disturbance agents (spruce budworm, spruce bark beetle) interposed upon a background of scattered smaller canopy gaps. Consequently, rates of disturbance fluctuated considerably over time. Reconstructed canopy gaps were temporally and spatially scattered; during disturbance peaks, they were both larger and more numerous.

Conclusions: Despite peaks in disturbance, several of which created relatively large gaps, this system has experienced no significant change in species composition. Instead, the shade-tolerant Picea rubens has maintained canopy dominance. The patch dynamics described here consist of dramatic structural, not compositional, changes to the forest. The persistence of Picea rubens is attributed to a combination of traits: (1) abundance of advance regeneration; (2) ability to endure suppression and respond favourably to release; and (3) longevity relative to ecologically similar species.

Nomenclature: Gleason & Cronquist (1991) for vascular plants; Crum (1983) for mosses.

Abbreviation: DWD = Down woody debris.