Milan Chytrý

Determination of diagnostic species with statistical fidelity measures

Abstract Statistical measures of fidelity, i.e. the concentration of species occurrences in vegetation units, are reviewed and compared. The focus is on measures suitable for categorical data which are based on observed species frequencies within a vegetation unit compared with the frequencies expected under random distribution. Particular attention is paid to Bruelheides u value. It is shown that its original form, based on binomial distribution, is an asymmetric measure of fidelity of a species to a vegetation unit which tends to assign comparatively high fidelity values to rare species. Here, a hypergeometric form of u is introduced which is a symmetric measure of the joint fidelity of species to a vegetation unit and vice versa. It is also shown that another form of the binomial u value may be defined which measures the asymmetric fidelity of a vegetation unit to a species. These u values are compared with phi coefficient, chi‐square, G statistic and Fishers exact test. Contrary to the other measures, phi coefficient is independent of the number of relevés in the data set, and like the hypergeometric form of u and the chi‐square it is little affected by the relative size of the vegetation unit. It is therefore particularly useful when comparing species fidelity values among differently sized data sets and vegetation units. However, unlike the other measures it does not measure any statistical significance and may produce unreliable results for small vegetation units and small data sets. The above measures, all based on the comparison of observed/expected frequencies, are compared with the categorical form of the Dufrêne‐Legendre Indicator Value Index, an index strongly underweighting the fidelity of rare species. These fidelity measures are applied to a data set of 15 989 relevés of Czech herbaceous vegetation. In a small subset of this data set which simulates a phytosociological table, we demonstrate that traditional table analysis fails to determine diagnostic species of general validity in different habitats and large areas. On the other hand, we show that fidelity calculations used in conjunction with large data sets can replace expert knowledge in the determination of generally valid diagnostic species. Averaging positive fidelity values for all species within a vegetation unit is a useful approach to measure quality of delimination of the vegetation unit. We propose a new way of ordering species in synoptic species‐by‐relevé tables, using fidelity calculations. Nomenclature:Ehrendorfer (1973).

Separating habitat invasibility by alien plants from the actual level of invasion.

Habitats vary considerably in the level of invasion (number or proportion of alien plant species they contain), which depends on local habitat properties, propagule pressure, and climate. To determine the invasibility (susceptibility to invasions) of different habitats, it is necessary to factor out the effects of any confounding variables such as propagule pressure and climate on the level of invasion. We used 20 468 vegetation plots from 32 habitats in the Czech Republic to compare the invasibility of different habitats. Using regression trees, the proportion of alien plants, including archaeophytes (prehistoric to medieval invaders) and neophytes (recent invaders), was related to variables representing habitat properties, propagule pressure, and climate. The propagule pressure was expressed as the proportion of surrounding urban and industrial or agricultural land, human population density, distance from a river, and history of human colonization in the region. Urban and industrial land use had a positive effect on the proportion of both archaeophytes and neophytes. Agricultural land use, higher population density, and longer history of human impact positively affected the proportion of archaeophytes. Disturbed human-made habitats with herbaceous vegetation were most invaded by both groups of aliens. Neophytes were also relatively common in disturbed woody vegetation, such as broad-leaved plantations, forest clearings, and riverine scrub. These habitats also had the highest proportion of aliens after removing the effect of propagule pressure and climate, indicating that they are not only the most invaded, but also most invasible. These habitats experience recurrent disturbances and are rich, at least temporarily, in available nutrients, which supports the hypothesis that fluctuating resources are the major cause of habitat invasibility. The least invaded habitats were mires and alpine-subalpine grasslands and scrub. After removing the effect of propagule pressure and climate, some habitats actually invaded at an intermediate level had very low proportions of aliens. This indicates that these habitats (e.g., dry, wet, and saline grasslands, base-rich fens, and broad-leaved deciduous woodlands) are resistant to invasion.

Statistical determination of diagnostic species for site groups of unequal size

Abstract Aim: Concentration of species occurrences in groups of classified sites can be quantified with statistical measures of fidelity, which can be used for the determination of diagnostic species. However, for most available measures fidelity depends on the number of sites within individual groups. As the classified data sets typically contain site groups of unequal size, such measures do not enable a comparison of numerical fidelity values of species between different site groups. We therefore propose a new method of measuring fidelity with presence/absence data after equalization of the size of the site groups. We compare the properties of this new method with other measures of statistical fidelity, in particular with the Dufrêne-Legendre Indicator Value (IndVal) index. Methods: The size of site groups in the data set is equalized, while relative frequencies of species occurrence within and outside of these groups are kept constant. Then fidelity is calculated using the phi coefficient of association. Results: Fidelity values after equalization are independent of site group size, but their numerical values vary independently of the statistical significance of fidelity. By changing the size of the target site group relative to the size of the entire data set, the fidelity measure can be made more sensitive to either common or rare species. We show that there are two modifications of the IndVal index for presence/absence data, one of which is also independent of the size of site groups. Conclusion: The phi coefficient applied to site groups of equalized size has advantages over other statistical measures of fidelity based on presence/absence data. Its properties are close to an intuitive understanding of fidelity and diagnostic species in vegetation science. Statistical significance can be checked by calculation of another fidelity measure that is a function of statistical significance, or by direct calculation of the probability of observed species concentrations by Fishers exact test. An advantage of the new method over IndVal is its ability to distinguish between positive and negative fidelity. One can also weight the relative importance of common and rare species by changing the equalized size of the site groups. Nomenclature: Ehrendorfer (1973).

Plot sizes used for phytosociological sampling of European vegetation

Abstract In European phytosociology, variable plot sizes are traditionally used for sampling different vegetation types. This practice may generate problems in current vegetation or habitat survey projects based on large data sets, which include relevés made by many authors at different times. In order to determine the extent of variation in plot sizes used in European phytosociology, we collected a data set of 41 174 relevés with an indication of plot size, published in six major European journals focusing on phytosociology from 1970 to 2000. As an additional data set, we took 27 365 relevés from the Czech National Phytosociological Database. From each data set, we calculated basic statistical figures for plot sizes used to sample vegetation of various phytosociological classes. The results show that in Europe the traditionally used size of vegetation plots is roughly proportional to vegetation height; however, there is a large variation in plot size, both within and among vegetation classes. The effect of variable plot sizes on vegetation analysis and classification is not sufficiently known, but use of standardized plot sizes would be desirable in future projects of vegetation or habitat survey. Based on our analysis, we suggest four plot sizes as possible standards. They are 4 m2 for sampling aquatic vegetation and low-grown herbaceous vegetation, 16 m2 for most grassland, heathland and other herbaceous or low-scrub vegetation types, 50 m2 for scrub, and 200 m2 for woodlands. It has been pointed out that in some situations, sampling in either small or large plots may result in assignment of relevés to different phytosociological classes or habitat types. Therefore defining vegetation and habitat types as scale-dependent concepts is needed.

Weed vegetation of arable land in Central Europe: Gradients of diversity and species composition

Abstract Question: What are the main broad-scale spatial and temporal gradients in species composition of arable weed communities and what are their underlying environmental variables? Location: Czech Republic and Slovakia. Methods: A selection of 2653 geographically stratified relevés sampled between 1954–2003 was analysed with direct and indirect ordination, regression analysis and analysis of beta diversity. Results: Major changes in weed species composition were associated with a complex gradient of increasing altitude and precipitation and decreasing temperature and base status of the soils. The proportion of hemicryptophytes increased, therophytes and alien species decreased, species richness increased and beta diversity decreased with increasing altitude. The second most important gradient of weed species composition was associated with seasonal changes, resulting in striking differences between weed communities developed in spring and summer. In summer, weed communities tended to have more neophytes, higher species richness and higher beta diversity. The third gradient reflected long-term changes in weed vegetation over past decades. The proportion of hemicryptophytes and neophytes increased, while therophytes and archaeophytes decreased, as did species richness over time. The fourth gradient was due to crop plants. Cultures whose management involves less disturbances, such as cereals, harboured less geophytes and neophytes, and had higher species richness but lower beta diversity than frequently disturbed cultures, such as root crops. Conclusions: Species composition of Central European weed vegetation is mainly influenced by broad-scale climatic and edaphic factors, but its variations due to seasonal dynamics and long-term changes in agricultural management are also striking. Crop plants and crop-specific management affect it to a lesser, but still significant extent. Nomenclature: Kubát et al. (2002).

LOCAL AND REGIONAL PATTERNS OF SPECIES RICHNESS IN CENTRAL EUROPEAN VEGETATION TYPES ALONG THE pH/CALCIUM GRADIENT

We investigated the relationship between soil pH/calcium content and species richness of vascular plants in seven broadly defined Central European vegetation types, using Ellenberg indicator values for soil reaction and a phytosociological data set of 11,041 vegetation sample plots from the Czech Republic. The vegetation types included (A) broad-leaved deciduous forests, (B) meadows, (C) dry grasslands, (D) reed-bed and tall-sedge vegetation, (E) fens and transitional mires, (F) perennial synanthropic vegetation and (G) annual synanthropic vegetation. Relationships between local species richness (alpha diversity) and pH/calcium were positive for vegetation types A and C, negative for D and G, unimodal for E, and insignificant for B and F. Ellenberg soil reaction values explained 37% of variation in local species richness for vegetation type E, 24% for A, 13% for D, but only less than 4% for the others. Species pool size, i.e., the number of species that can potentially occur in a given habitat, was calculated for each plot using Beals index of sociological favourability applied to a large phytosociological database. For most vegetation types, the relationships between species pool size and pH/calcium were similar to the relationships between local species richness and pH/calcium, with the exception of meadows (weak unimodal) and perennial synanthropic vegetation (weak negative).These patterns suggest that for those types of Central European vegetation that developed independently of human influence in the Pleistocene or early Holocene (dry grasslands, deciduous forests), there are larger pools of calcicole than calcifuge species. This pattern is also found at the level of local species richness, where it is, however, less clearly pronounced, possibly due to the predominance of a few widespread and generalist calcifuges in acidic habitats. The unimodal pattern found in mires may result from similar underlying mechanisms, but in high pH environments mineral-rich spring waters probably decrease species richness by having toxic effects on plant growth. By contrast, vegetation types developed under direct human influence (meadows, synathropic vegetation) show weak negative or no relationships of local species richness or species pool to pH/calcium gradient. These results support the hypothesis ofPärtel (Ecology 83: 2361–2366, 2002) andEwald (Folia Geobot. 38: 357–366, 2003), that the modern calcicole/calcifuge disparity in the species pool of Central European flora has resulted from historical and evolutionary processes that took place on high pH soils. In the Pleistocene, calcareous soils dominated both the dry continental landscapes of Central Europe and glacial refugia of temperate flora, which were mostly situated in southern European mountain ranges with abundant limestone and dolomite. The negative pattern of species richness along the pH/calcium gradient found in reed-bed and tall-sedge vegetation, however, is not consistent with this historical explanation.

ALIEN PLANTS IN TEMPERATE WEED COMMUNITIES: PREHISTORIC AND RECENT INVADERS OCCUPY DIFFERENT HABITATS

Variables determining the number of native and alien plants on arable land in Central Europe are identified. Species richness of 698 samples of weed floras recorded in the Czech Republic in plots of a standard size of 100 m 2 in 1955-2000 was studied in relation to altitudinally based floristic region, soil type, type of cultivated crop, climatic variables, altitude, year of the record, crop cover and height, and human population density in the region. Vascular plant species were classified into native and alien, the latter divided in archaeophytes, introduced before AD 1500, and neophytes, introduced after this date. The use of minimal adequate models in the analysis of covariance allowed determination of the net effects of mutually correlated environmental variables. Models for particular species groups explained 33-48% of variation in species numbers and 27-51% in propor- tions; however, explanatory variables affected native species, archaeophytes, and neophytes differently. The number and proportion of neophytes increased in 1955-2000, whereas the number of native species and archaeophytes declined (in archaeophytes more slowly in the warm than in the moderate to cool altitudinal floristic region). In warm and dry regions and on dry soils, where most archaeophytes find optimum conditions, fewer native species are able to persist in weed communities than in colder and wetter regions. Archaeophytes respond like neophytes to some variables (climate, seasonal development of crop) and alternatively like native species to other variables (increasing agricultural intensification through time, human population density). Archaeophytes are common in old crops intro- duced with the beginning of agriculture (cereals), but are poorly represented in relatively recently introduced crops (rape, maize), where neophytes are most numerous. These patterns reflect the history of plant invasions in Central Europe. Neolithic agriculture, introduced from the Near East in the sixth millenium BC, brought archaeophytes with crops and, by creating intense and continuous propagule pressure and imposing new agricultural man- agement, facilitated their invasion. By contrast, the crops introduced during the past five centuries and their specific agrotechnical management have supported spreading of other weed species, mainly invaders from overseas.

Stratified resampling of phytosociological databases: some strategies for obtaining more representative data sets for classification studies

Abstract Question: The heterogeneous origin of the data in large phytosociological databases may seriously influence the results of their analysis. Therefore we propose some strategies for stratified resampling of such databases, which may improve the representativeness of the data. We also explore the effects of different resampling options on vegetation classification. Methods: We used 6050 plot samples (relevés) of mesic grasslands from the Czech Republic. We stratified this database using (1) geographical stratification in a grid; (2) habitat stratification created by an overlay of digital maps in GIS; (3) habitat stratification with strata defined by traditional phytosociological associations; (4) habitat stratification by numerical classification and (5) habitat stratification by Ellenberg indicator values. Each time we resampled the database, taking equal numbers of relevés per stratum. We then carried out cluster analyses for the resampled data sets and compared the resulting classifications using a newly developed procedure. Results: Random resampling of the initial data set and geographically stratified resampling resulted in similar classifications. By contrast, classifications of the resampled data sets that were based on habitat stratifications (2–5) differed from each other and from the initial data set. Stratification 2 resulted in classifications that strongly reflected environmental factors with a coarse grain of spatial heterogeneity (e.g. macroclimate), whereas stratification 5 resulted in classifications emphasizing fine-grained factors (e.g. soil nutrient status). Stratification 3 led to the most deviating results, possibly due to the subjective nature of the traditional phytosociological classifications. Conclusions: Stratified resampling may increase the representativeness of phytosociological data sets, but different types of stratification may result in different classifications. No single resampling strategy is optimal or superior: the appropriate stratification method must be selected according to the objectives of specific studies. Abbreviations: ASS = Phytosociological association; ELL = Ellenberg indicator values; GEO = Geographical stratification; GIS = Geographical information system; NUM = Numerical classification; RAN = Random resampling.

Trends in species diversity and composition of urban vegetation over three decades

Abstract Question: What was the change in diversity of urban synantropic vegetation in a medium-sized Central European city during the period of increasing urbanization (1960s–1990s)? Location: The city of Plzeň, an industrial centre of the western part of the Czech Republic. Methods: Sampling of various types of synanthropic vegetation, conducted in the 1960s, was repeated by using the same methods in the 1990s. This yielded 959 relevés, of which 623 were made in the 1960s and 336 in the 1990s. The relevés were assigned to the following phytosociological classes: Chenopodietea, Artemisietea vulgaris, Galio-Urticetea, Agropyretea repentis and Plantaginetea majoris. Total number of vascular plant species, evenness index J, number of alien species (classified into archaeophytes and neophytes), and mean Ellenberg indicator values for light, temperature, continentality, moisture, soil reaction, and nutrients were obtained for each relevé. Results: From 1960s to 1990s, there was a significant decrease of species richness and diversity in synanthropic vegetation. The proportion of archaeophytes decreased in most vegetation types, indicating the contribution of this group of species, often confined to specific rural-like habitats, to the observed impoverishment of ruderal vegetation. The proportion of neophytes did not change between the two periods. Comparison between 1960s and 1990s indicated a decrease in light, temperature, moisture, soil reaction and nutrient indicator values in some vegetation types. In both periods, Artemisieta, Galio-Urticetea and Chenopodietea formed a distinct group harbouring more species than Agropyretea and Plantaginetea. Neophytes, i.e. recently introduced species, were most represented in the early successional annual vegetation of Chenopodietea, rather than in perennial vegetation of the other classes. Conclusions: Synanthropic vegetation of Plzeň exhibited a general trend of decrease in species diversity. Nomenclature: Ehrendorfer (1973) for species; Moravec et al. (1995) for phytosociological units.

Towards unification of national vegetation classifications: a comparison of two methods for analysis of large data sets.

In European phytosociology, national classifications of corresponding vegetation types show considerable differences even between neighbouring countries. Therefore, the European Vegetation Survey project urgently needs numerical classification methods for large datasets that are able to produce compatible classifications using datasets from different countries. We tested the ability of two methods, TWINSPAN and COCKTAIL, to produce similar classifications of wet meadows (Calthion, incl. Filipendulenion) for Germany (7909 releves) and the Czech Republic (1287 releves) in this respect. In TWINSPAN, the indicator ordination option was used for classification of two national datasets, and the extracted assignment criteria (indicator species) were applied crosswise from one to the other national dataset. Although the datasets presumably contained similar community types, TWINSPAN revealed almost no correspondence between the groups derived from the proper classification of the national dataset and the groups defined by the assignment criteria taken from the other national dataset. The reason is probably the difference in structure between the national datasets, which is a typical, but hardly avoidable, feature of any pair of phytosociological datasets. As a result, the first axis of correspondence analysis, and consequently the first TWINSPAN division, are associated with different environmental gradients; the difference in the first division is transferred and multiplied further down the hierarchy. COCKTAIL is a method which produces releve groups on the basis of statistically formed species groups. The user determines the starting points for the formation of species groups, and groups already found in one dataset can be tested for existence in the other dataset. The correspondence between the national classifications produced by COCKTAIL was fairly good. For some releve groups, the lack of correspondence to groups in the other national dataset could be explained by absence of the corresponding vegetation types in one of the countries, rather than by methodological problems.