Long‐term effects of the abandonment of grazing on steppe‐like grasslands

Question: What are the consequences of grazing abandonment on the Stipa lessingiana dominated steppe‐like grasslands? What is the relative importance of management and environmental factors in causing variation in species composition and abundance in the continuously grazed and abandoned grassland stands? Location: Transylvanian Lowland, Romania. Methods: Repeated vegetation mapping of a grassland stand, where grazing was abandoned 35 years ago; re‐sampling six grassland stands surveyed 29–57 years ago. For revealing long‐term changes in species composition and rank abundance PCoA ordination was applied. The relative importance of management and environmental factors in structuring vegetation were explored by CCA ordination. Diversity, evenness and the relative number and abundance of red‐listed species were compared between managed and abandoned stands. Results: Our results pointed out that grasslands which were formerly grazed and dominated by S. lessingiana, in the long‐term absence of grazing, have been transformed into a S. pulcherrima dominated type. Management, probably by creating bare surfaces and preventing litter accumulation, had the strongest effect on the species composition and abundance in the grasslands. Abandoned grassland stands had lower diversity and evenness compared to continuously grazed stands. While at the same time, the relative number of threatened, rare species did not differ between managed and abandoned sites. Conclusion: Maintaining extensively grazed, as well as un‐managed, Stipa dominated grasslands would be important in order to create various habitat conditions for plant species, especially threatened and rare species, and promote diversity on the landscape scale.     

Within‐stand spatial structure and relation of boreal canopy and understorey vegetation

Question: How do spatial patterns and associations of canopy and understorey vegetation vary with spatial scale along a gradient of canopy composition in boreal mixed‐wood forests, from younger Aspen stands dominated by Populus tremuloides and P. balsamifera to older Mixed and Conifer stands dominated by Picea glauca? Do canopy evergreen conifers and broad‐leaved deciduous trees differ in their spatial relationships with understorey vegetation? Location: EMEND experimental site, Alberta, Canada. Methods: Canopy and understorey vegetation were sampled in 28 transects of 100 contiguous 0.5 m × 0.5 m quadrats in three forest stand types. Vegetation spatial patterns and relationships were analysed using wavelets. Results: Boreal mixed‐wood canopy and understorey vegetation are patchily distributed at a range of small spatial scales. The scale of canopy and understorey spatial patterns generally increased with increasing conifer presence in the canopy. Associations between canopy and understorey were highly variable among stand types, transects and spatial scales. Understorey vascular plant cover was generally positively associated with canopy deciduous tree cover and negatively associated with canopy conifer tree cover at spatial scales from 5–15 m. Understorey non‐vascular plant cover and community composition were more variable in their relationships with canopy cover, showing both positive and negative associations at a range of spatial scales. Conclusions: The spatial structure and relation of boreal mixed‐wood canopy and understorey vegetation varied with spatial scale. Differences in understorey spatial structure among stand types were consistent with a nucleation model of patch dynamics during succession in boreal mixed‐wood forests.     

Vegetation mapping using hierarchical object‐based image analysis applied to aerial imagery and lidar data

Aims: The primary objective of this study is to map the distribution and quantify the cover of vegetation alliances over the entirety of San Clemente Island (SCI). To this end, we develop and evaluate the mapping method of hierarchical object‐based classification with a rule‐based expert system. Location: San Clemente Island, California, USA. Methods: We developed and tested an approach based on hierarchical object‐based classification with a rule‐based expert system to effectively map vegetation communities on SCI following the Manual of California Vegetation classification system. In this mapping approach, the shrub species defining each vegetation community and non‐shrub growth forms were first mapped using aerial imagery and lidar data, then used as input in an automated mapping rule set that incorporates the percent cover rules of a field‐based mapping rule set. Results: The final vegetation map portrays the distribution of 19 vegetation communities across SCI, with the largest areas comprised of California Annual and Perennial Grassland (35%) and three types of coastal sage scrub and maritime succulent scrub, comprising a combined 53% of the area. Map accuracy was assessed to be 79% based on fuzzy methods and 61% with a traditional accuracy assessment. The accuracy of tree identification was assessed to be 81%, but species‐level tree accuracy was 45%. Conclusions: Semi‐automated approaches to vegetation community mapping can produce repeatable maps over large spatial extents that facilitate ecological management efforts. However, some low‐statured shrub community types were difficult to differentiate due to patchy canopies of co‐occurring species including abundant non‐native grasses characteristic of complex disturbance histories. Species‐level tree mapping accuracy was low due to the difficulty of identifying species within poorly illuminated canyons, resulting from sub‐optimal image acquisition timing.     

Understorey vegetation development in North Finnish Picea forests after disturbance: re‐analysis of Sirén's data

Abstract. Sirén (1955) studied understorey species composition, tree stand properties and humus‐layer thickness in 64 unlogged forest stands on topographically and pedologically comparable sites. The stands were of even age (6 – 300 yr), stocked with the first or second tree generation after wildfire. The view of Sirén and several authors after him, that the vegetation of old‐growth boreal Picea forests is homogeneous on a broad scale, was examined by applying, in parallel, the partial variants of two ordination methods (DCA and PCA) to Sirén's vegetation data. Two main vegetation gradients were found: a major gradient running from recently burnt plots with prominence of pioneer species to plots with stand age > 100 yr, a well stocked tree layer and a thick humus layer, dominance of feather‐mosses and ample occurrence of shade‐tolerant as well as light‐preferring vascular plant species, and a second gradient along which first‐ and second‐generation plots segregate. The more prominent element of Betula trees in first‐ than in second‐generation stands < 100 yr contributed to the latter. A minor third gradient related to humus‐layer thickness was recovered by partial DCA only. The main vegetation gradient reappeared in separate ordinations of data from 47 mature forest stands (> 100 yr), but without being correlated with forest age. Variation among mature‐forest stands in the importance of pioneer species is considered mainly to be brought about by fine‐scale disturbance processes such as tree uprooting. Increasing importance of factors operating on within‐stand scales [development of a varied gap structure and stronger gradients in tree influence (radiation at ground level), soil moisture, soil depth and nutrient availability] with time is also reflected in the second and third mature‐forest ordination axes. Possible implications of the results for conservation of biological diversity and monitoring of changes in boreal forests are discussed.     

Mapping continuous fields of forest alpha and beta diversity

Question: How to map continuous fields of forest alpha and beta diversity in remote areas, based on easily accessible spatial data. Location: Kyrgyzstan/Central Asia. Methods: The study relied on a combination of predictive mapping and remote sensing. Punctual measurements of alpha diversity were linked to topography and reflectance using regression models. For beta diversity, ordination techniques were employed to extract major vegetation gradients. Scores on the ordination axes were regressed against topography as well as reflectance and subsequently mapped. Beta diversity was mapped as spatial turnover rate along these axes. Results: The diversity maps quantified species counts and turnover in a spatially contiguous manner while taking into account fuzzy transitions. The variance explained by regression models ranged from 51% to 61% in cross‐validation. Many of the observed differences were caused by differences in species shares. The occurrence of walnut, in particular, showed a negative relation to woody species numbers. Conclusion: Mapping biodiversity in remote areas can be based on easily accessible spatial data in combination with a set of calibration field samples. With regard to human influence on walnut dominance, a total removal of human land use would be counterproductive in terms of diversity conservation. The results of this study highlight the need for comprehensive analyses of diversity patterns that include spatially contiguous quantifications of species numbers, shares and turnover rates.     

Aspen succession and nitrogen loading: a case for epiphytic lichens as bioindicators in the Rocky Mountains,USA

Question: Can lichen communities be used to assess short‐ and long‐term factors affecting seral quaking aspen (Populus tremuloides) communities at the landscape scale? Location: Bear River Range, within the Rocky Mountains, in northern Utah and southern Idaho, USA. Method: Forty‐seven randomly selected mid‐elevation aspen stands were sampled for lichens and stand conditions. Plots were characterized according to tree species cover, basal area, stand age, bole scarring, tree damage, and presence of lichen species. We also recorded ammonia emissions with passive sensors at 25 urban and agricultural sites throughout an adjacent populated valley upwind of the forest stands. Nonmetric multidimensional scaling (NMS) ordination was used to evaluate an array of 20 variables suspected to influence lichen communities. Results: In NMS, forest succession explained most variance in lichen composition and abundance, although atmospheric nitrogen from local agricultural and urban sources also significantly influenced the lichen communities. Abundance of nitrophilous lichen species decreased with distance from peak ammonia sources and the urban center in all aspen succession classes. One lichen, Phaeophyscia nigricans, was found to be an effective bioindicator of nitrogen loading. Conclusions: Lichen communities in this landscape assessment of aspen forests showed clear responses to long‐term (stand succession) and short‐term (nitrogen deposition) influences. At the same time, several environmental factors (e.g. tree damage and scarring, distance to valley, topography, and stand age) had little influence on these same lichen communities. We recommend further use of epiphytic lichens as bioindicators of dynamic forest conditions.     

Insect community response to switchgrass intercropping and stand age of loblolly pine (Pinus taeda) plantations

相似文献   

Understorey vegetation change in a Picea mariana chronosequence

Eighteen black spruce (Picea mariana) stands, representing postfire ages of 26 to 120 yr, were surveyed for understorey vegetation and site/microsite characteristics at two spatial scales. This enabled comparison of within- versus among-stand compositional variation.Detrended correspondence analysis (DCA) ordination among the 18 stands revealed a complex age/moisture gradient. DCA ordination among 1 800 quadrats within the stands indicated a similar gradient with much compositional overlap. Quadrats were grouped, using two-way indicator species analysis (TWINSPAN), into 9 classes each representing a phase in understorey vegetation composition. These phases shifted in abundance from young to old stands with a high degree of concordance among replicates in the same age class. Understorey succession is strongly linked to the stages in tree growth, mortality and thinning coupled with the accumulation of site moisture.Abbreviations DCA Detrended Corrospondence Analysis     

Composition Patterns of Caribbean Limestone Forests: Are Parsimony,Classification, and Ordination Analyses Congruent?1

This study presents an analysis of floristic composition patterns for limestone vegetation from three Caribbean islands: Hispaniola, Mona, and Puerto Rico. The physical setting of these communities varies from very dry to wet climates, and from ridgetops, slopes, to plateaus. Consequently, vegetation communities have variable species composition. The questions addressed were: (1) What are the patterns of species composition among limestone vegetation types based on a parsimony analysis of species assemblages (PASA) and how congruent are they widi geography, climate, topography, and vegetation physiognomy? and (2) How do PASA patterns compare to floristic patterns obtained with a classification (UPGMA) and an ordination (NMS)? The main distinction of floristic categories was between communities of humid and dry climates, followed by a finer separation of communities congruent with topography; vegetation physiognomy corresponded with both. PASA, UPGMA, and NMS produced highly coincident floristic groups; however, affinities among groups were different. Advantages of PASA over UPGMA and NMS are that PASA produces groupings strictly based on the shared species and provides different measures of support for floristic groups. The three techniques indicated that the humid and dry limestone areas of Puerto Rico had floristically divergent dry‐type vegetation, even when they were structurally and physiognomically convergent. Also, floristic affinities of dry communities in Puerto Rico were stronger with dry communities on other islands than to more humid communities on the same island; thus, humidity regime is stronger than geography in promoting floristic links among limestone vegetation communities of die Caribbean. Almost every conclusion obtained from UPGMA and NMS was also taken from PASA, and so all diree techniques are compatible. The methodological, theoretical, and interpretive simplicity of PASA is what makes it an attractive procedure for studies that examine composition patterns.     

The direct regeneration hypothesis in northern forests

Question: Can the direct regeneration hypothesis (DRH) be used to predict post‐disturbance regeneration after fire, wind disturbance, and clearcutting in northern forests? Do life‐history traits such as regeneration strategy and shade tolerance influence post‐disturbance regeneration success of tree species? Location: Northern forests in North America. Methods: A meta‐analysis was conducted by collecting published data on pre‐ and post‐disturbance stand compositional characteristics in the northern forests. For each tree species, compositional difference (CD) was calculated as the difference between basal area proportions of the post‐ and pre‐disturbance stands, but for post‐disturbance stands <25 years of age, post‐disturbance proportions were calculated based on relative stem density. Results: Species response to disturbances was best explained by regeneration strategy, while disturbance type had no effect on CD. The proportion of broadleaf trees with either strong or weak vegetative reproduction ability increased after all disturbances. Serotinous species had CD values not significantly different from zero after fire, while CD for semi‐serotinous species was negative. The post‐disturbance proportions of non‐serotinous conifers decreased after all forms of disturbance. Conclusions: All disturbances promote broadleaf trees, regardless of regeneration strategy (suckering, sprouting, or seeding). The DRH is supported for conifers with serotinous cones after fire. Fire causes local extinction of non‐serotinous conifers, while wind and clearcutting only decrease the proportion of non‐serotinous conifers because of partial survival of seed sources and advanced regeneration. This study suggests that increasing stand‐replacing disturbances associated with global climate change will promote broadleaf trees in northern forests.     

Restoration of natural vegetation in degraded Imperata cylindrica grassland: understorey development in forest plantations

Abstract. Reclamation of former, degraded forest lands occupied by Imperata cylindrica is one of the crucial environmental and forestry issues in the humid tropics, notably Southeast Asia. We suggest that it is possible to gradually restore the original natural forest cover with the help of a sacrifice fallow crop of fast-growing exotic tree species. Recently, a set of suitable fast-growing plantation tree species has been identified and stand establishment methods developed for this purpose. We assessed the regeneration of natural vegetation in stands of different plantation tree species and evaluated the ecological impact of species composition in the plantation understorey. PCA ordination, regression analysis and analysis of covariance were applied at different stages of the study. We found a marked vegetational resemblance between stands dominated by Acacia mangium: they had the highest number of indigenous trees in their understorey, whereas stands of other plantation trees supported more diverse grass and herb vegetation. A high proportion of evergreen woody vegetation reduces the risk of fire and grass competition and enhances secondary succession towards natural forest.     

Understory vegetation indicates historic fire regimes in ponderosa pine‐dominated ecosystems in the Colorado Front Range

Question: Can current understory vegetation composition across an elevation gradient of Pinus ponderosa‐dominated forests be used to identify areas that, prior to 20th century fire suppression, were characterized by different fire frequencies and severities (i.e., historic fire regimes)? Location: P. ponderosa‐dominated forests in the montane zone of the northern Colorado Front Range, Boulder and Larimer Counties, Colorado, USA. Methods: Understory species composition and stand characteristics were sampled at 43 sites with previously determined fire histories. Indicator species analyses and indirect ordination were used to determine: (1) if stands within a particular historic fire regime had similar understory compositions, and (2) if understory vegetation was associated with the same environmental gradients that influence fire regime. Classification and regression tree analysis was used to ascertain which species could predict fire regimes. Results: Indicator species analysis identified 34 understory species as significant indicators of three distinct historic fire regimes along an elevation gradient from low‐ to high‐elevation P. ponderosa forests. A predictive model derived from a classification tree identified five species as reliable predictors of fire regime. Conclusions: P. ponderosa‐dominated forests shaped by three distinct historic fire regimes have significantly different floristic composition, and current understory compositions can be used as reliable indicators of historical differences in past fire frequency and severity. The feasibility demonstrated in the current study using current understory vegetation properties to detect different historic fire regimes, should be examined in other fire‐prone forest ecosystems.     

The Quantitative Analysis of Subtropic Forests in Hwang-shan

1.The subtropic secondary forests of Mt. Hwangshan in the province Anhui, have been studied by both the "point-centred Quarter method" and a quantitative analysis which was based on the continuum concept of vegetation. 2. Each stand sampled was both by a 20 random points sampling and 16 m two ×l6 m quardrat sampling. It seemed that the random points sampling method gave a more realistic representation of the stands as judged by the floristic data and the relative abundence and relative dominance data (table 1 and 2). 3. The Jaceard's "Coefficient of community", Sorenson's "Quotient of similarity", Whittaker's "Percentage of similarity" "X2 and" test of fit of goodless of fit" of Greig-Smith were was to measure the community similarity (or homogenity) in the same stand. Our modified "Percentage of similarity" with weighing by relative abundance of all the tree species seems better when compared with the other criteria (table 3). 4. An index called "Importance value similarity" (denoted as c), modified from the "Coefficient of community" of Curtis, had been designed for ordinating the stands. The ordination of the 18 stands of the present study with this index was shown in table 4. Ordination by another method, called the "Paper strips method", was shown in table 5. The phenomenon of vegetation continua was clearly shown in both kinds of ordination.     

Stand characteristics,understorey associates and environmental correlates ofEucalyptus tetrodonta F. Muell. forests on Gunn Point,northern Australia

An ordination study of 20Eucalyptus tetrodonta forest stands growing on deep earths in monsoonal Australia revealed two major gradients in understorey vegetation type. The first axis reflected both floristic and structural understorey variation, where litter and shrub cover were inversely related to grass cover. This axis is thought to reflect a complex fire-vegetation type interaction, where vegetation is primarily determined by the saturation of the soil profile in the wet season, as measured by the colour of the iron rich soils. On the second axis of the ordination, floristic composition but not vegetation structure, and stand height were found to vary with the intercorrelated measures of soil gravel and moisture supply.E. tetrodonta is able to regenerate in the absence of fire, but firing appears to stimulate regeneration. All stands contained some advance growth, which occurs in distinct clumps, probably reflecting these plants clonal origin. Sapling presence in the stands is variable and the recruitment of advance growth into this size class appears to be related to over-wood competition. The size class distribution of trees was found to be similar amongst the stands, therefore stand structure appeared to be independent of understorey type. In comparison to general models of temperate eucalypt regeneration processes the tropical eucalypts have evolved different regeneration strategies, possibly in response to the severe annual drought.Nomenclature: Chippendale (1971), unless otherwise indicated. Structural classification: Specht (1970).I acknowledge the help of Mr. Clyde Dunlop who identified plant specimens and provided stimulating discussion, and Dr. Peter Minchin for permission to use the computer package ECOPACK to prepare and handle the floristic data, and his invaluable assistance with the ordination analysis. Mr. Lee Belbin and CSIRO Division of Water and Land Resources kindly gave their permission to use the Numerical Taxonomy Package (NTP).     

Interim Responses of Littoral River Channel Vegetation to Reestablished Flow after Phase I of the Kissimmee River Restoration Project

Measurements of littoral vegetation stands and species‐level surveys of associated plant communities were made in channels of the Kissimmee River from 1998 through 2008, a period that spanned channelized, non‐flowing conditions through 7 years of near‐continuous reestablished flow. Dissected by flood control canal C‐38 in 1971, the river was virtually without flow until early 2001, when Phase I of the Kissimmee River Restoration Project (KRRP) reestablished flow to a central section of river channel. This study evaluated the effects of reestablished flow on littoral vegetation in river channels as an indicator of system status and progress toward the project goal of ecological integrity. Predictions of vegetation response to reestablished flow included reduction in the width of vegetation stands, and changes in the growth‐form composition of littoral stands from near‐equal dominance by floating and emergent species to overwhelming dominance by emergent growth forms. Variables included plant cover by species and growth‐form, width of vegetation stands, and vegetated percentage of channel. Under the currently incomplete (interim) status of the KRRP, results for littoral vegetation stands indicate trends in the predicted directions of change, and three of four predicted changes have occurred. Vegetation stand widths decreased substantially and littoral plant communities became heavily dominated by emergent species; BACIPS (before‐after‐control‐impact‐paired series) analyses indicated significant restoration effects for most littoral stand metrics.     

Overstory vegetation and successional trends of Land Between The Lakes,USA

133 stands from the uplands of Land Between The Lakes (LBL) were sampled. Stand data were ordinated to identify clinal patterns in the overstory vegetation. A coeno-cline, accounting for 52% of the species variance, was developed from the first ordination axis. Classification of stands identified nine dominance types: Pinus echinata-Quercus spp., Q. prinus, Q. prinus-Q. alba, Q. stellata-Q. alba, Q. velutina-Q. alba, Q. alba, Q. alba-Q. rubra, Acer saccharum-Q. alba-Fagus grandifolia sadF. grandifolia-A. saccharum. Compositionally-stable (climax) and unstable (successional) stands were analyzed separately. Trajectories of unstable stands include the succession of P. echinata to Quercus-dominated communities on xeric sites and of Quercus types to more mesophytic communities (e.g. A. saccharum and F. grandifolia) onmesic sites. These shifts indicate the nature of the future climax overstory composition at LBL. The first axis DCCA ordination for climax stands accounted for 45% of the vegetation variance. Factor Analysis identified two significant suites of variables related to vegetation: soil chemistry/soil moisture and soil texture. Specific variables related to variation in the climax vegetation include soil pH, calcium and magnesium, effective soil depth, available water capacity, percent rock in the soil, slope position, aspect, elevation and distance to the opposing slope. Modal positions of species along the direct gradient ordination were similar in rank order to those derived from the indirect gradient ordination of all stands. The latter ordination represents the present forest vegetation pattern at LBL. Climax stands showed a shift in modes of Quercus spp. toward the xeric end of the gradient and increases in habitat width of mesophytic species. Following a reduction in disturbance, species appear to adjust to changes in competition and available habitat. That is, the dimensions of realized niches (i.e. habitat widths) change with succession. The climax vegetation at LBL appears to show centrifugal organization.     

Broad‐scale vegetation‐environment relationships in Eurasian high‐latitude areas

Question: How is tundra vegetation related to climatic, soil chemical, geological variables and grazing across a very large section of the Eurasian arctic area? We were particularly interested in broad‐scale vegetation‐environment relationships and how well do the patterns conform to climate‐vegetation schemes. Material and Methods: We sampled vegetation in 1132 plots from 16 sites from different parts of the Eurasian tundra. Clustering and ordination techniques were used for analysing compositional patterns. Vegetation‐environment relationships were analysed by fitting of environmental vectors and smooth surfaces onto non‐metric multidimensional scaling scattergrams. Results: Dominant vegetation differentiation was associated with a complex set of environmental variables. A general trend differentiated cold and continental areas from relatively warm and weakly continental areas, and several soil chemical and physical variables were associated with this broad‐scaled differentiation. Especially soil chemical variables related to soil acidity (pH, Ca) showed linear relationships with the dominant vegetation gradient. This was closely related to increasing cryoperturbation, decreasing precipitation and cooler conditions. Remarkable differences among relatively adjacent sites suggest that local factors such as geological properties and lemming grazing may strongly drive vegetation differentiation. Conclusions: Vegetation differentiation in tundra areas conforms to a major ecocline underlain by a complex set of environmental gradients, where precipitation, thermal conditions and soil chemical and physical processes are coupled. However, local factors such as bedrock conditions and lemming grazing may cause marked deviations from the general climate‐vegetation models. Overall, soil chemical factors (pH, Ca) turned out to have linear relationship with the broad‐scale differentiation of arctic vegetation.     

Response of pine natural regeneration to small‐scale spatial variation in a managed Mediterranean mountain forest

Questions: What influence do management practices and previous tree and shrub stand structure have on the occurrence and development of natural regeneration of Pinus sylvestris in Mediterranean mountain forests? How are the fine‐scale and environmental patterns of resources affected and what impact does this have on the distribution of the regeneration? Location: A Pinus sylvestris Mediterranean mountain forest in central Spain. Methods: Upperstory trees and regeneration (seedlings and saplings) were mapped in four 0.5‐ha plots located in two types of stand with different management intensities (even‐aged and uneven‐aged stands). Environmental variables were recorded at the nodes of a grid within the plots. The relationships between the upperstory and regeneration were evaluated by bivariate point pattern analysis; redundancy analysis ordination and variation partitioning were performed to characterize regeneration niches and the importance of the spatial component. Results: Seedlings and saplings presented a clumped structure under both types of management and their distribution was found to be related to the spatial distribution of favourable microsites. Regeneration was positively related to conditions of partial cover with high soil water content during the summer. More than half of the explained variance was spatially structured in both types of stand. This percentage was particularly high in the even‐aged stands where the pattern of regeneration was highly influenced by the gaps created by harvesting. Conclusions: The spatial distribution of the tree and shrub upperstory strongly influences regeneration patterns of P. sylvestris. Current management practices, promoting small gaps, partial canopy cover and moderate shade in even‐aged stands, or favouring tree and shrub cover in the case of uneven‐aged stands, appears to provide suitable conditions for the natural regeneration of P. sylvestris in a Mediterranean climate.     

Vegetation differentiation in the patterned landscape of the central Everglades: importance of local and landscape drivers

Aim We present a model to account for self‐assembly of the slough–ridge–tree island patterned landscape of the central Everglades in southern Florida via feedbacks among landforms, hydrology, vegetation and biogeochemistry. We test aspects of this model by analysing vegetation composition in relation to local and landscape‐level drivers. Location We quantified vegetation composition and environmental characteristics in central Water Conservation Area (WCA) 3A, southern WCA‐3A and southern WCA‐3B in southern Florida, based on their divergence in water management and flow regimes over the past 50 years. Methods In 562 quadrats, we estimated species coverages and quantified maximum, minimum and average water depth, soil depth to bedrock, normalized difference vegetation index (NDVI) and proximity to the nearest tree island. We used non‐metric multi‐dimensional scaling (NMS) to relate compositional variation to local and landscape‐level factors, and evaluated environmental differences among eight a priori vegetation types via anova . Results Water depth and hydroperiod decreased from sloughs to ridges to tree islands, but regions also differed significantly in the abundance of several community types and the hydroregimes characterizing them. NMS revealed two significant axes of compositional variation, tied to local gradients of water depth and correlated factors, and to a landscape‐scale gradient of proximity to tall tree islands. Sawgrass height and soil thickness increased toward higher ridges, and NDVI was greatest on tree islands. Main conclusions This study supports four components of our model: positive feedback of local substrate height on itself, woody plant invasion and subsequent P transport and concentration by top predators nesting on taller tree islands, compositional shifts in sites close to tree islands due to nutrient leakage, and flow‐induced feedback against total raised area. Regional divergence in the relationship of community types to current hydroregimes appears to reflect a lag of a few years after shifts in water management; a longer lag would be expected for shifts in landscape patterning. Both local and landscape‐level drivers appear to shape vegetation composition and soil thickness in the central Everglades.