Ordination Diagram Research Articles

Biomonitoring for effective management of wastewater discharges and the health of the river environment

The management of wastewater effluents in South Africa has to date been based on physico-chemical standards although recently the assimilative capacity of the receiving stream has been included but not widely implemented. Very few assessments of the success of these management practices in protecting the downstream aquatic environment have been done. Bioassessments using benthic macro invertebrates were conducted up and down stream of a number of wastewater works to measure the degree of impact that the works were having on the stream ecosystems. A significant deterioration in the benthic invertebrate fauna downstream of most of the wastewater works was observed. This was particularly noticeable where the receiving stream was of good quality while there was little change in poor quality receiving streams. This was confirmed using ordination techniques. The ordination diagrams displayed how community composition varied with important environmental variables although some of these may be surrogates for other unmeasured variables. It was revealed that river flow and the dilution of the effluent were the most important of the measured variables explaining the variability in the biota. Escherichia coli and soluble phosphorus were also shown to be important with suspended solids also significant. Most of the wastewater works, in the month prior to the biological investigation, had complied with their permit requirements. The exceptions, found on infrequent occasions at different works, were soluble phosphorus, chlorine, E. coli and conductivity. One permit requirement that was consistently ignored was a statement forming part of the effluent standards permit system which states that the effluent should not contain constituents that are poisonous or injurious to animal (including aquatic) life. No systems for monitoring this criterion were in place. Despite the general compliance with regulations governing wastewater treatment a lack of adequate effluent dilution appeared to be the major cause of degradation of the downstream biota. These wastewater works are found in a region that does not have an abundant supply of water for dilution so treatment of the effluent to a condition that is not injurious to aquatic life becomes a necessity. Thus it appears that the physico-chemical standards used to characterise and manage wastewater treatment in South Africa are inadequate to protect the downstream environment in the vicinity below many wastewater works.

97/02535 A question concerning coal structure and/or reactivity. The reaction of maleic anhydride with Pittsburgh No. 8 coal

The chapter reviews the main measures of association between objects or descriptors that are used in ecology to cluster objects or descriptors, or produce ordination diagrams in spaces of reduced dimensionality. The chapter includes discussion of the following topics: Q and R analyses (three dimensional data box, O-mode of analysis, P-mode of analysis, Q-mode of analysis, R-mode of analysis, S-mode of analysis, T-mode of analysis, choosing between Q and R mode, A space, and metric or Euclidean space), association coefficients (similarity coefficients, distance coefficients, dependence coefficients, double-zero problem, asymmetrical coefficient, and symmetrical coefficient), similarity coefficients in Q mode (symmetrical binary similarity coefficients, asymmetrical binary similarity coefficients, symmetrical quantitative similarity coefficients, missing values, partial similarity matrix, asymmetrical quantitative similarity coefficients, and probabilistic similarity coefficients), distance coefficients in Q mode (metric properties, dissimilarity coefficients, metric distance coefficients, and semimetric distance coefficients), coefficients of dependence in R mode (coefficients of dependence for descriptors other than species abundances, coefficients of dependence for species abundances, and biological associations), choice of a coefficient (criteria, and ordered comparison case series), and transformations for community composition data (transformation formulas, chord transformation, species profile transformation, Hellinger transformation, chi-square distance transformation, chi-square metric transformation, Beals smoothing). Numerical methods are illustrated with real ecological applications, drawn from the literature. The chapter ends on a description of relevant software implemented in the R language; it also cites some commercially available statistical packages and programs from researchers.

REMNANT FOREST STANDS AT A PRAIRIE ECOTONE SITE: PRESETTLEMENT HISTORY AND COMPARISON WITH OTHER MAPLE-BASSWOOD STANDS

The maple-basswood community type has long been associated with the “Big Woods” of Minnesota and adjacent Wisconsin, although this community type also exists in discontinuous phases within surrounding forest types. This study looks at the apparent most northwestern outlier of the maple-basswood community type, Rydell National Wildlife Refuge (NWR) in Minnesota. Specific goals are (1) to determine if the maple-basswood stands at Rydell NWR existed historically, and (2) to compare the composition of the Rydell NWR stands to maple-basswood “core” sites, as well as other outlier sites. Public Land Survey records for the period between 1872 and 1876 for a 16-township block encompassing Rydell NWR indicate that, although this area was dominated by an oak-aspen forest type, a distinct maple-basswood region existed prior to European settlement and this had escaped disturbance from fire and wind. Present-day stand composition data from six Rydell stands are compared with published data from other maple-basswood stands, using detrended correspondence analysis. A strong geographic pattern is indicated in the ordination diagram, which is attributed to differences in September precipitation and actual evapotranspiration between the sites. Local-scale environmental gradients act to modify the dominant climatic trend on composition, however, as some Rydell stands plotted closely to the core region in ordination space, whereas other Rydell stands demonstrated greater similarity to oak and aspen stands in North Dakota. [Key words: forest-prairie ecotone, maple-basswood, presettlement vegetation, ordination, Minnesota.]

Outbreak of multi-drug resistant Staphylococcus aureus: A cluster analysis

Background. Outbreaks of multi-drug resistant Staphylococcus aureus over eight years were investigated to prevent future endemic. Methods. Isolates of multiresistant S. aureus underwent a cluster analysis combined with canonical discriminant analysis using bacteriologic biotyping and sensitivity to 21 drugs. Results. Of a total of 786 strains recovered from 155 in-patients, the specialty surgical ward (SW) exhibited 470 isolates (59.8%) and the general SW 214 (27.2%). Among six clusters formed, four clusters were predominant in the general SW. An ordination diagram from the canonical discriminant analysis revealed a distribution in which clusters were localized temporally (year) and spatially (ward). A yearly shift of clusters indicated emergence of a new phenotype of multiresistant S. aureus. Conclusions. The cluster analysis of isolates of multiresistant S. aureus using biotyping and sensitivity may supplement the classical method of tracing the spreading patterns of this microbe.

Temporal and Spatial Variation in the Demography of Gentiana Pneumonanthe, a Rare Perennial Herb

1 A total of 35 transition and elasticity matrices for the rare iteroparous herb Gentiana pneumonanthe was analysed for temporal and spatial variation. The data used were collected annually from 10 sites in six different populations on up to seven occasions in the period 1987-93. 2 In general, temporal variation was higher in transition matrices than in elasticities, while between-site variation was high for both transition and elasticity matrices. 3 The relative contributions of three life-history transitions, progression (or growth, G), recruitment from seed (fecundity, F) and survival (retrogression plus stasis, L) to the finite rate of increase, A, were also highly variable, between years within as well as between sites. The observed variation is very large in comparison with that previously observed either between or within other iteroparous herbs from open habitats, even showing some overlap with demographic patterns normally characteristic of woody plants. 4 A G-L-F ordination shows a long, narrow band across the entire diagram from matrices with a low elasticity for L and high elasticities for G and F at one end to matrices which have an elasticity of 1.00 for L on the other. 5 Correlations between the G-, Land F-elasticities, A, and the vegetation cover suggest that the band in the ordination diagram represents a successional pathway in wet heathlands from invasive to regressive populations. Elasticity matrices from regularly mown hay meadows are characteristic of stable populations, but represent senile, regressive populations after mowing has ceased. 6 Relationships between A and the elasticities of G, L and F show that in matrices of stable or declining populations (A 1) the contribution of progression and fecundity becomes larger. 7 Given their relatively large temporal and spatial variation, elasticities are useful in nature conservation and management only if the corresponding value of A is taken into account. 8 Significant positive correlations between transition probabilities of different life stages were observed. This phenomenon may increase the risk of extinction by environmental stochasticity.

Detrended correspondence analysis in the ordination of data for phenetics and cladistics

SummaryParnell, J. & Waldren, S.: Detrended correspondence analysis in the ordination of data for phenetics and cladistics. – Taxon 45: 71‐84. 1996. – ISSN 0040‐0262.Most methods of morphometric and phylogenetic analysis do not permit the incorporation of both categorical and quantitative data. Detrended correspondence analysis (DCA) was used to produce ordination diagrams of strictly categorical data from a cladistic analysis, and mixed data from a morphometric study of hybridization. In both cases it proved preferable to standardize the data prior to DCA, because increased separation on ordination was found when larger scale ranges were used. A standardization is suggested, such that characters are scaled on a range from 1 to the total number of characters scored; this will produce more informative ordinations if large numbers of characters are scored. When used in conjunction with appropriate cladograms this phenetic technique may help reveal features such as homoplasy which are not readily apparent in cladograms, while the capability of DCA to combine both categorical and quantitative characters is especially useful in morphometric studies and in areas of evolutionary biology, e.g. chronistic studies of extinction events in morphospace.

Canonical correspondence analysis and related multivariate methods in aquatic ecology

Canonical correspondence analysis (CCA) is a multivariate method to elucidate the relationships between biological assemblages of species and their environment. The method is designed to extract synthetic environmental gradients from ecological data-sets. The gradients are the basis for succinctly describing and visualizing the differential habitat preferences (niches) of taxa via an ordination diagram. Linear multivariate methods for relating two set of variables, such as twoblock Partial Least Squares (PLS2), canonical correlation analysis and redundancy analysis, are less suited for this purpose because habitat preferences are often unimodal functions of habitat variables. After pointing out the key assumptions underlying CCA, the paper focuses on the interpretation of CCA ordination diagrams. Subsequently, some advanced uses, such as ranking environmental variables in importance and the statistical testing of effects are illustrated on a typical macroinvertebrate data-set. The paper closes with comparisons with correspondence analysis, discriminant analysis, PLS2 and co-inertia analysis. In an appendix a new method, named CCA-PLS, is proposed that combines the strong features of CCA and PLS2.

Long-term vegetation recovery after vehicle track abandonment on Dartmoor, SW England, U.K.

The use of extensive track networks on northern Dartmoor has caused considerable damage to vegetation during the past 24 years. The number and use of tracks has increased over this period. This paper describes the natural recovery of vegetation on tracks abandoned at several times in the past on five different sites. The undamaged vegetation varies from Agrostis grassland to wet moorland dominated by Calluna vulgaris and/or Molinia caerulea. Analysis of environmental parameters shows that the most recently used tracks suffer compaction and reduced soil depth and moisture. Canonical Correspondence Analysis (CCA) is used to explore these effects further, and the way in which vegetation can be expected to change over time is deduced from comparison of tracks of different “ages” as a “time-track” on an ordination diagram. Recovery may be rapid in grassland and heathland sites but areas of higher moorland and blanket bog recover more slowly and may never revert to their original vegetation composition. The results are discussed with respect to practical conservation management of the area.

Aquatic macrophytes and ecological gradients in 57 small lakes in southern Finland

Records made in the late 1970s in 57 small (0.5–200 ha) lakes in southern Finland (about 61°30′N) were used to examine ecological gradients, species assemblages, and two a priori classifications of lakes. The a priori groupings were: (1) general limnological types (hypertrophic, eutrophic, clear-water meso-eutrophic, mesotrophic, clear-water oligotrophic and brown-water oligotrophic lakes); (2) the traditional Finnish botanical lake type classification. The material was classified with TWIN-SPAN and ordinated with detrended correspondence analysis (DCA) and canonical correspondence analysis (CCA). The species richness and the occurrence of various growth forms vary greatly, being highest in the clear-water meso-eutrophic and eutrophic lakes, and lowest in brown-water oligotrophic waters. For helophytes, however, species richness was highest in eutrophic and hypertrophic lakes. The trophic state (indicated by the specific conductivity and the summer water pH) determines the occurrence of many helophyte species; for hydrophytes the principal determinants are trophic state and water transparency. There was a relatively good correspondence between the results of multivariate analyses and a priori classifications of lakes, especially in most nutrient-poor and nutrient-rich, as well as in some clear-water lakes. The mesotrophic and eutrophic lake categories, do, however, overlap widely. The ordination diagrams indicate an ecological continuum, and only relative importance of lake classifications. Ninety-four macrophyte species separate into a few main groups along the oligotrophy-eutrophy gradient.

The relationship of chrysophycean stomatocysts to environmental variables in freshwater lakes in British Columbia

The relationships between the distributions of 82 chrysophycean cyst morphotypes and measured environmental variables in freshwater lakes in British Columbia were examined using ordination and regression statistics. After removal of unusual samples, 60 lakes were included in the analyses. Indirect and direct gradient analysis explained 23.2 and 14.0% of the variance in the cyst distribution data, and 31.4 and 53.7% of the variance in the cyst–environment relationship, respectively. Watershed area, Secchi depth and [Mg] were identified as the variables with the greatest contributions to the first ordination axis. Maximum depth contributed most strongly to axis 2. Constrained redundancy analyses were used to test the ability of individual environmental variables to explain the variance in the cyst data; no one variable was shown to have an overriding effect on cyst distributions. Five groups of cysts were identified using the ordination diagrams and the correlations between each cyst and each environmental variable. Partial least squares regression was used to construct inference models that quantified the relationship between the cyst distributions and four environmental variables (pH, [Mg], total phosphorus, and Secchi depth). For each variable, the best model included only those cysts which were significantly correlated with that variable. The inference model for pH yielded the strongest relationship (r2 = 0.51) and best predictive ability (root mean square error of prediction = 0.32). All the inference models showed a strong trend in the residuals, such that inferences at the low end of the observed gradient tended to be overestimates and inferences at the high end tended to be underestimates. Thus, paleolimnological inferences of past environmental conditions using these models will tend to underestimate the degree of change. Key words: British Columbia, phytoplankton, Chrysophyceae, stomatocysts, paleolimnology, eutrophication.

Environmental factors affecting submerged macrophyte communitites in regulated waterbodies in Egypt

Regulated waterbodies developed to generate hydroelectric power have a number of features which differentiate them from natural waterbodies. The submerged aquatic vegetation of three regulated Egyptian waterbodies (Lake Nasser, Aswan Reservoir and the River Nile at Aswan) is described in relation to 22 environmental physico-chemical determinands. The three waterbodies are interconnected but show major ditferences in water level regime. These are produced by the requirements of operating schedules for hydroelectric power generation. Macrophyte growth, distribution and community composition were directly related to environmental variables using Detrended Canonical Correspondence Analysis (DCCA). The resulting ordination diagram shows that variation in aquatic vegetation can best be explained by: (1) hydrosoil Ca 2+ , Mg 2+ and organic matter, (2) water NO 3 − ; and (3) elevation datum. The implications of these results for assessment of environmental impact on aquatic macrophytes in regulated waterbodies are discussed

Grazing in remnant woodland vegetation: changes in species composition and life form groups

Abstract. Grazing by domestic livestock in native woodlands can have major effects on ecosystem functioning by the removal of plant species that form important functional groups. This paper documents the changes in floristics in a large group of remnants of native woodland left after agricultural clearing in southwestern Australia. Species richness and diversity were significantly reduced in remnants and the proportion of exotic species increased. Detrended Correspondence Analysis (DCA) was used to identify floristic and environmental patterns among plots and identified two distinct groups based on grazing intensity. This indicated that the significance of the relationship between grazing effects and DCA floristic axes was greater than edaphic characteristics that normally influence floristic patterns. Floristic characteristics of sites that were influencing the position of plots on the ordination diagram included proportion of exotic species and proportion of native perennial shrubs and herbs. Numbers of species of native shrubs and perennial herbs were significantly reduced in grazed plots and numbers of exotic annual grasses and herbs were significantly higher. Other life form groups such as native perennial grasses and geophytes were not significantly affected by grazing. Reproductive strategies of perennial species showed a significant decrease in numbers of resprouters and a significant increase in numbers of facultative seeder/sprouters. Exclosure plots showed increases in number and cover of perennial shrubs and herbs after three years whereas number and cover of exotic species did not change. Time series DCA showed that the floristic composition of exclosure plots in grazed sites became closer to that of the ungrazed sites.

Relationships between macroinvertebrate communities and environmental variables in metal‐contaminated streams in south‐west England

SUMMARY1. Macroinvertebrate species and information on thirty‐nine variables were recorded at forty‐six sites on twelve Cornish streams affected to varying extents by past metalliferous mining.2. Relationships between macroinvertebrate communities and environmental variables were examined using canonical correspondence analysis (CCA). Copper was the strongest correlate with Axis 1 of the analysis, suggesting that it may have a major role in determining community structure. There were also strong correlations between Axis 1 and aluminium, alkalinity, pH, dissolved organic matter and algal cover, and between Axis 2 and discharge.3. CCA distinguished four site groupings. In spite of seasonal changes in position on the ordination, as revealed through CCA and detrended correspondence analysis (DCA), sites retained their group membership. There were differences in the proportions of the major taxa within the four CCA‐derived groups. In the group with the highest copper concentration, Ephemeroptera were absent, but triclads and chironomids were abundant, with Orthocladiinae the dominant group.4. Changes in the occurrence of species along the metal gradient reflected changes in sensitivity. Tolerance' plots of selected species based on the CCA site ordination diagram were used to identify environmental thresholds.5. As well as direct toxic effects of copper and aluminium on invertebrate communities, co‐precipitation of these two metals could be an important physical factor. Further modifying influences resulting from interactions between toxic metals and pH, alkalinity, hardness and dissolved organic matter contributed to the complexity of conditions affecting the faunal community in metal‐contaminated streams.6. The most severely contaminated sites, with mean copper concentration exceeding 500 μgl−1, were all characterized by a reduced community dominated by the flatworm Phagocata vitta, the chironomids Chaetodadius melaleucus and Eukiefferiella claripennis and the net‐spinning caddis Plectrocnemia conspersa. The necessity for species identification is demonstrated in relation to variation in metal tolerance.

On the statistical analysis of vegetation change: a wetland affected by water extraction and soil acidification

Abstract. A case study is presented on the statistical analysis and interpretation of vegetation change in a wetland subjected to water extraction and acidification, without precise information on the environmental changes. The vegetation is a Junco‐Molinion grassland and the changes in vegetation are evaluated on the basis of relevés in 1977 and 1988 of 20 plots in a small nature reserve on moist oligotrophic, Pleistocene sands in the Netherlands. The changes are attributed to water extraction (since 1972) and soil acidification and the effect of the environmental changes on the vegetation is inferred from data on water depth and acidity collected in 1988.Many species typical of wetlands decreased in abundance, including rare species such as Parnassia palustris, Selinum carvifolia and Ophioglossum vulgatum. Some species increased, notably Anthoxanthum odoratum, Holcus lanatus and Plantago lanceolata. A significant decrease was found in the mean Ellenberg indicator values for moisture and acidity. The mean indicator value for nutrients did not change significantly.Multivariate analysis of the species data by Redundancy Analysis demonstrated the overall significance of the change in species composition between 1977 and 1988 (P < 0.01, Monte Carlo permutation). The spatial and temporal variation in the species data was displayed in ordination diagrams and interpreted in terms of water depth and pH.A simple model is developed to infer the change in water depth and pH from the relevé data and recent data on water depth and pH. Because the correlation between water depth and pH made a joint estimation of the changes useless, the change in pH was estimated for a series of likely changes in water depth. For the most likely change in water depth, significant acidification was inferred from the change in vegetation. The model is more generally applicable as a constrained calibration method.

Autecological response of Kalmia angustifolia to forest types and disturbance regimes

Vegetative and reproductive growth of Kalmia angustifolia was studied in eight habitats which differed in vegetation cover, soil type, moisture and disturbance. The shrub was most vigorous, with the highest cover (73%), plant height (55 cm), current year's twig density (283 m −2) and aboveground biomass (1139 gm −2), in the open canopy black spruce forest. Under the closed canopy undisturbed black spruce forest, Kalmia remained mostly vegetative with only 1 twig m −2 producing flowers, whereas in the clear-cut and burned forest site Kalmia flowered profusely (67 flowering twigs m −2). Mean of individual leaf area of older Kalmia growing under a closed canopy black spruce forest was significantly higher (6.88 cm 2) than younger plants developing from sprouts after clearcutting and burning (2.35 cm 2). In the clear-cut and burned site Kalmia produced the highest density of fruits and seeds (806 fruits and 60 630 seeds m −2, respectively) compared with the other sites (10–451 fruits and 1824–1841 592 seeds m −2, respectively). However, more than 50% of the seeds were non-viable. On the other hand, although very few seeds were produced under the closed canopy forest more than 80% of them were viable seeds. Discriminant analysis (DA) of 14 vegetative parameters of Kalmia, measured in the eight sites, classified all the sites that were disturbed by cutting and/or burning in one group irrespective of their soil type, moisture regime and time since disturbance. The undisturbed sites were positioned separately in the DA ordination diagram based on their forest cover type and soil moisture regime. Disturbance-induced spread of Kalmia and its implications for forest management are discussed.

Factors controlling the distribution of vascular and non-vascular epiphytes in the northern Andes

The epiphytic vegetation on host trees along an altitudinal transect on the west slope of the Central Cordillera in Colombia was sampled. Bark-type restricted sampling comprised four full-grown forest trees, not necessarily four species, at altitudinal intervals of c. 200 meters from 1000 to 4130 m a.s.l. and included the canopy. Both vascular and non-vascular epiphytes, the latter often more abundant in the montane rain forest, were included in the analyses. Using a method of direct gradient analysis, canonical correspondence analysis, the variation explained by various environmental variables could be discriminated from a great amount of variation that seemed not related to any ecological factor. To a large extent, the randomness in propagule supply appears to determine the floristic composition on branch/trunk segments. The grouping of releves in a phytosociological analysis concurred with a clustering of samples in an ordination diagram of the first two extracted constrained axes. The sample scores on these two axes were strongly correlated with the complex variables ‘altitude’ and ‘height within the host tree’. Specialists and generalists with respect to the two variables were defined based on tolerances provided by the canonical correspondence analysis as was the position of species on the gradients involved. Independently from any of the other variables entered, a relation between the epiphytic vegetation and host species was detected, particularly in the case of Brunellia occidentalis, a fast growing tree species of higher altitude. No relation between chemical characteristics of suspended soil in the Upper Montane Rain forest and its supporting species could be demonstrated.

Canonical community ordination. Part I: Basic theory and linear methods

AbstractCanonical community ordination comprises a collection of methods that relate species assemblages to their environment, in both observational studies and designed experiments. Canonical ordination differs from ordination sensu stricto in that species and environment data are analyzed simultaneously. Part I reviews the theory in a non-mathematical way with emphasis on new insights for the interpretation of ordination diagrams. The interpretation depends on the ordination method used to create the diagram. After the basic theory, Part I is focused on the ordination diagrams in linear methods of canonical community ordination, in particular principal component analysis, redundancy analysis and canonical correlation analysis. Special attention is devoted to the display of qualitative environmental variables.

Canopy arthropods on Scots pine: influence of season and stand age on community structure and the position of sawflies (Diprionidae) in the community

The response of arthropods to season and age of stands was studied in the crowns of managed, pure Scots pine stands ( Pinus sylvestris L.) by insecticide fogging. Pine monocultures were characterised by a stable guild composition (defoliators being predominant) rather than by the composition of taxonomic groups (orders) which are subject to seasonality (mainly Hymenoptera, Coleoptera, Lepidoptera, Homoptera). The second variable (age) had more influence on the community composition than the first (season). In total, significant differences related to the age of the pines were found among most of the arthropod taxa, the medium-aged stands showing a higher density of arthropods in the crown strata. The evaluation of abundance of Hymenoptera (sawfly larvae being the majority in samples—eight species present) by canonical correspondence analysis, in relation to the two environmental variables examined, showed their shift towards the spring (June) in the ordination diagram, when the population density of pine sawfly larvae was higher in comparison with the late summer period. The age variable was not significant in diprionid response. Quantitative data on the most numerous arthropods are included.

Weed infestation and factors affecting weed incidence in spring cereals in Finland - a multivariate approach

Weed vegetation of spring cereal fields in southern and central Finland was analyzed by ordination methods to provide a community level description of weed populations. Attention was paid particularly to the relative importance of environmental factors affecting weed incidence such as crop management, soil properties and weather conditions. A data set of 33 weed taxa from 252 fields was subjected to both indirect and direct gradient analysis. Indirect ordination was obtained with correspondence analysis (CA), and direct gradient analyses were performed with redundancy analysis (RDA) and with canonical correspondence analysis (CCA) relating environmental factors to the occurrence of weeds. Among several management factors, continuous herbicide use explained best the variation in the species composition of weed flora. Weed vegetation was also associated with soil type, moisture conditions and soil pHH2O. Ordination diagrams visualized the species-environment interactions and detected characteristic weed species for different geographical regions. In addition to ordination analyses of weed flora, the level and structure of weed infestation are described. The density of weeds averaged 170 plants m2 (median=124) and the air-dry weight of weeds 320 kg ha-1 (median=183). The average weed density was the same in different soil types, but the weed biomass was lower in clay soils than in coarse mineral and organic soils

Habitat selection of breeding and migrating icterine warblers Hippolais icterina: a multivariate study

Site occupancy of migrating and breeding lcterine warblers was studied with respect to unoccupied randomly chosen sites in a willow and poplar mixed forest along the river Danube in Hungary Multivariate discriminant analysis revealed 4 different zones in the ordination diagram exclusive breeding zone, exclusive migrating zone overlapping zone, and an unused zone They were separated according to foliage cover and tree height Tree species composition mostly affects habitat use through physiognomy There was a partially overlapping, but clear difference in habitat selection between the migrating and the breeding birds