Global Ecology and Biogeography, (Global Ecol. Biogeogr.) (2005) 14, 263–270 Blackwell Publishing, Ltd.

RESEARCH PAPER

Ecological behaviour of herbaceous forest species along a pH gradient: a comparison between oceanic and semicontinental regions in northern France Christophe Coudun* and Jean-Claude Gégout

LERFoB, UMR INRA-ENGREF 1092, Ecole Nationale du Génie Rural, des Eaux et des Forêts, 14, Rue Girardet, CS 4216, 54042 Nancy Cedex, France

ABSTRACT

Aim On the basis of 2402 phytoecological relevés, with complete species lists, and real pH measurements resulting from chemical analyses of the top layer of forest soils, this paper compares quantitatively the ecological response of 46 herbaceous forest species along a pH gradient in two regions of northern France. Location The two regions investigated are oceanic north-western France (NW) and semicontinental north-eastern France (NE). Methods For each of the 46 species with more than 50 occurrences in both NW and NE regions, an ecological response curve was computed with simple logistic regression models, and two synthetic numerical values were derived: ecological optimum (OPT) and ecological amplitude (AMP). A comparison of the ecological behaviour of species present in both regions was performed in terms of shift in optimum and/or amplitude. Results All 46 species did reveal a reaction to pH. Our main observation was the relative stability of the behaviour of most species with reference to pH conditions, which is consistent with results from some previous studies. Slight differences regarding the two synthetic parameters (OPT and AMP) were however observed between the NW and NE regions, probably due to a form of competitive release in neutral environments in the NW. Main conclusions Among the main possibilities that could explain a geographical shift in the ecological behaviour of herbaceous species, competitive effects are, we suggest, the most plausible explanation. In the light of former studies, it seems that further analyses of geographical shifts in the ecological behaviour of forest plant species over large areas are needed, such as for example over the European continent.

*Correspondence: Christophe Coudun, LERFoB, UMR INRA-ENGREF 1092, Ecole Nationale du Génie Rural, des Eaux et des Forêts, 14, Rue Girardet, CS 4216, 54042 Nancy Cedex, France. E-mail: [email protected]

Nomenclature Tutin et al. (2001). Keywords Competitive release, ecological amplitude, ecological optimum, ecological response curves, EcoPlant database, Ellenberg’s indicator values, forest herbs, France, logistic regression.

Ecological niche modelling for plant species has received much attention in the last few decades in order to quantify the response of species to abiotic factors. Many statistical techniques have been developed to draw and interpret response curves, allowing a deeper understanding of species–environment relationships (Guisan & Zimmermann, 2000), with generalized linear modelling (GLM) or generalized additive modelling (GAM)

techniques being most intensively used (e.g. Austin et al., 1984; Odland et al., 1995). Recent international meetings have led to several publications on species distribution modelling and prediction, representing both theoretical and practical developments (Guisan et al., 2002; Lehmann et al., 2002; Scott et al., 2002). The establishment of unimodal response curves allows the definition of indicator values that are of interest to synthesize ecological information on species (ter Braak & Looman, 1986; Lawesson & Oksanen, 2002). Ecological optimum and amplitude

© 2005 Blackwell Publishing Ltd www.blackwellpublishing.com/geb

DOI: 10.1111/j.1466-822X.2005.00144.x

INTRODUCTION

263

C. Coudun and J.-C. Gégout are traditionally derived from computations (ter Braak & Gremmen, 1987; Odland et al., 1995) and facilitate the comparison of reactions to ecological factors among species. Some empirical indicator values have been proposed by Ellenberg et al. (1992) for central continental Europe, which have been used with success in many other locations (Diekmann, 2003). However, Ellenberg’s indicator values were revisited in some studies to account for local conditions, such as in the Netherlands (Ertsen et al., 1998), Denmark (Lawesson & Mark, 2000), Sweden (Diekmann, 1995) or oceanic Great Britain (Hill et al., 1999, 2000). Besides varying with space, the ecological indicator value of a particular plant species may also change with time during the life of the plant (Parrish & Bazzaz, 1985). Since the beginning of the 1990s, many studies have been conducted on the comparison of ecological response of plants in different regions (Thompson et al., 1993; Diekmann & Lawesson, 1999; Hill et al., 2000; Prinzing et al., 2002; Gégout & Krizova, 2003). Diekmann & Lawesson (1999) have suggested three main explanations for a possible shift in ecological behaviour: (i) species may differentiate themselves into ecotypes adapted to conditions in different regions (Turesson, 1922); (ii) the Walters’ rule, also known as the habitat-constancy rule (Walter & Walter, 1953), might force species to compensate an ecological factor by another one and lead to changes in pH optimum or amplitude of species; or (iii) the presence/absence of competitors may cause a shift in the indicator values. In France, empirical indicator characteristics for forest plant species were derived by Rameau et al. (1989, 1993) and in the early 1970s, Le Tacon & Timbal (1972) had noticed some differences in the ecological behaviour of some forest plant species between beech stands in north-eastern and northwestern France, with regard to the humus status of the soil. Diekmann & Lawesson (1999) noticed that pH was one of the main underlying variables determining the floristic variation within forests in their four study regions and also that field observations or experiments were ‘largely lacking’. Too often, the value of environmental factors is determined from weighted averages of species indicator values (e.g. Diekmann, 1996; Dzwonko, 2001). The pH represents a complex gradient, but it has been considered as a direct environmental gradient (Austin, 1980) because it has a direct physiological effect on plant growth. It can also be considered as a resource gradient since it controls plants’ uptake of minerals and is, at the same time, correlated to many edaphic and climatic factors (Duchaufour, 1989; Falkengren-Grerup et al., 1995; Tyler, 2003). In this study, our main objective was to investigate quantitatively the ecological response of forest plant species to pH in two regions of northern France (semicontinental north-eastern, NE, and oceanic north-western, NW), on the basis of real pH measurements resulting from chemical analyses of the top layer of forest soils. This constitutes a major difference with respect to previous studies that were based on computations with empirical Ellenberg’s indicator values. From a comparison between oceanic NW and semicontinental NE, we investigated whether forest herbaceous species reveal similar ecological response curves in rather similar ecological conditions but in different geographical locations. 264

MATERIALS AND METHODS Two regional data sets Two regional data sets were extracted from EcoPlant (Gégout, 2001), a forest phytoecological database now integrating more than 120 different relevés sources, scattered all over the French continental territory and the island of Corsica. The main objective of EcoPlant is to store complete species lists with all available associated environmental information (climate- and soil-related parameters) either measured in the field, chemically analysed in laboratories or derived from geographical information systems. EcoPlant is currently used to derive analytical ecological indicator values for forest plant species with regard to climatic and edaphic factors (Gégout, 2001). For the present study, 73 relevés sources were used (the reference list is available on request to the authors) and two regions in northern France were defined, based on a 300 km distance to seashore (north-western, NW and north-eastern, NE, see Fig. 1). We separated oceanic and semicontinental plots consistently with existing bioclimatic maps of France (Bessemoulin, 1989); then we had two similar regional data sets in terms of the number of relevés (1353 forest sites for NW and 1049 for NE). To keep all the major ecological components, except mineral nutrition (pH), as constant as possible, the effects of temperature were neutralized during the sampling part of the study. The mean yearly temperature thus lies in the 8.5–11.5 °C range and the altitude is below 600 m, for all sites. The effect of light was also implicitly neutralized because all relevés come from closed forest environments. Our data set covered the whole pH gradient in both regions, with pH values varying between 3.4 and 8.1. All pH values are pH (H2O) laboratory values measured for the upper organo-mineral A horizon of soils. We used a pragmatic separation of the pH range into two classes, to account for acidic (3.4 < pH < 5.5) and neutral to rich (5.5 < pH < 8.1) forest sites (e.g. Gough et al., 2000). The value of 5.5 for pH also represents a value under which toxicity of aluminium might be important for plants (Espiau & Peyronel, 1977; Badeau, 1998). Among the 1353 forest sites from the NW region, 1083 were acidic and 270 were neutral to rich according to the previous description, and among the 1049 forest sites from the NE region, 668 were acidic and 381 were neutral to rich. Most sampled forest sites (quadrats) presented an area of 400 m2, consistent with current phytoecological practice, and complete species lists were available for each site (presence/ absence information). In order to be integrated in further analyses, herbaceous forest species had to be present at least 50 times in each region (i.e. presence in at least 5% of the forest sites of the region). In total, the species pool investigated here represents 46 herbaceous species. Simple logistic regression model The ecological response of species to pH was derived from simple logistic regression models and interpretation of response curves (ter Braak & Looman, 1986). Logistic regression is a generalized

Global Ecology and Biogeography, 14, 263–270, © 2005 Blackwell Publishing Ltd

Ecological behaviour of herbaceous forest species along a pH gradient

Figure 1 Location of the 2402 investigated forest sites in the two regions of northern France (north-western, NW, and north-eastern, NE). The pH status is represented through symbol type.

linear modelling technique (McCullagh & Nelder, 1997), with a logit link function and binomial error distribution, and is often used to characterize species–environment relationships (Austin et al., 1984; ter Braak & Looman, 1986; Odland et al., 1995). We characterized the significance (at the 0.05 level) of the Gaussian logit model (bell-shaped unimodal response curve) against the linear logit model (increasing or decreasing sigmoidal response curve) or against the null model (no reaction and flat response curve) with a residual deviance test (McCullagh & Nelder, 1997) based on the Akaike Information Criterion (Akaike, 1973). All computations were performed with the S-Plus statistical package (MathSoft, 1999). The logistic regression modelling technique is, however, still in debate because it does not allow all possible kinds of shapes for the response curves (Austin et al., 1984; Austin, 1987, 2002; Austin & Nicholls, 1997; Oksanen, 1997; Lawesson & Oksanen, 2002). We avoided this problem by computing two parameters, which summarize the information of the response curves derived for each of the 46 herbaceous studied species: the ecological optimum (OPT) and the ecological amplitude (AMP). OPT is simply the pH value when the probability of presence reaches its maximum, and when this value fell within the 3.4 – 8.1 range (significant unimodal response), we could compute a confidence interval around it (ter Braak & Looman, 1986; Vetaas, 2000; Oksanen et al., 2001). AMP is the pH range in which presence conditions are optimal and was computed as the pH range

containing 80% of the distribution of probability of presence (Gégout & Pierrat, 1998). Such a measurement for amplitude allows a clear numerical comparison between regions and has the advantage of being applicable for any response shape, contrary to ter Braak & Looman’s (1986) measure of tolerance, which is only valid for Gaussian logit models. RESULTS General stability of the pH ecological response for most species In both regions, the 46 studied plant species did react to pH, justifying the choice of pH as an important ecological variable. Figure 2 shows some response curves for acidophilous Deschampsia flexuosa, neutrophilous Melica uniflora and basophilous Carex flacca, resulting from logistic regression modelling in the two regions. Optima and amplitudes of all 46 herbaceous species are presented in Fig. 3. We found that species tend to occupy the same pH niche positions in the NW and the NE: the coefficient of determination, R2, computed for the 46 species to characterize the link between the OPT values between NW and NE, is equal to 0.78, the coefficient of determination is equal to 0.57 for the AMP values between NW and NE. The proportion of apparent acidophilous (OPT < 5.5) and basophilous (OPT > 5.5) species in both regions is as follows:

Global Ecology and Biogeography, 14, 263– 270, © 2005 Blackwell Publishing Ltd

265

C. Coudun and J.-C. Gégout

Figure 2 The pH ecological response curves derived with logistic regression models in the NW and the NE for Deschampsia flexuosa (defl), Melica uniflora (meun) and Carex flacca (cafla). The points indicate the position of the pH optimum (OPT), and the thick lines indicate width of ecological amplitude (AMP).

most species (43 out of 46) belong to the same pH ecological group in both regions, with nine acidophilous species and 34 basophilous species. Only Rubus fruticosus, a species which might include different spectra of taxa in the two regions, was found to be acidophilous in the NW and basophilous in the NE. Athyrium filix-femina and Stellaria holostea were found to be basophilous in the NW and acidophilous in the NE. However, the ecological amplitude of all three mentioned species is relatively large. Competition for most favourable environments The density of species optima is maximal for pH values between 6 and 7, which represents a range in which mineral nutrition conditions are optimal in northern France’s forests. For pH values between 6 and 7, exchangeable aluminium toxicity does not occur any longer and toxicity due to active calcium carbonate (CaCO3) for higher pH values does not occur yet (see Fig. 4). In these pH conditions, the sum of exchangeable cations (calcium, magnesium and potassium) is also favourable, contributing to the optimal mineral nutrition conditions for which species might compete. There is a tendency for some species that are acid tolerant to neutrophilous in the NE to have a greater pH optimum in the NW than in the NE (e.g. Oxalis acetosella, Athyrium filix-femina, Galeopsis tetrahit, Melica uniflora or Geranium robertianum), while some species that are basophilous in the NE have a lower pH optimum in the NW (e.g. Carex flacca, Stachys sylvatica or Convallaria majalis, see Fig. 3a). All species revealing a shift in the optimum and/or amplitude greater than 0.5 units are listed in Table 1. In general, it appears that species in the NW have less extreme pH optima than in the NE (see Fig. 3a). The number of species per relevé is higher in the NE forests than in the NW forests for all pH values, especially for optimal values between 6 and 7 (see Fig. 5). Moreover, the regional pool of neutrophilous/basophilous species seems to be larger in the NE than in the NW. Indeed, 19 supplementary species could be found more than 50 times in the NE region, but not in the NW 266

Figure 3 Scatter diagrams of (a) pH optima (OPT) and (b) pH amplitudes (AMP) positions in the NW and the NE for the 46 forest herbaceous species for which an ecological curve was derived in both regions. See Table 1 for complete names of species.

region, and among them, 12 were basophilous. Among the 13 species found more than 50 times in the NW but not in the NE, only five species were basophilous. With fewer species per relevé on eutrophic soils and with a smaller pool of neutrophilous species in the NW region, we suggest that competitive effects are less prevalent in the NW than in the NE, and that this may explain the tendency towards optimal pH values in the NW. The inferred weaker competitive conditions in the NW could explain the shift of acid tolerant and basophilous species’ optima of some species in the NE to neutrophilous ones in the NW. In terms of amplitude, only acid tolerant species revealed a clear trend towards greater values in the NW (e.g. Melampyrum pratense, Oxalis acetosella or Dryopteris carthusiana), because they can adapt to a wide range of environments, from poor

Global Ecology and Biogeography, 14, 263–270, © 2005 Blackwell Publishing Ltd

Ecological behaviour of herbaceous forest species along a pH gradient

Figure 4 Average content in the first horizon of soil (A horizon) of exchangeable aluminium (Al, in meq per 100 g dry soil), the sum of nutritive exchangeable cations (S, in meq per 100 g dry soil) and active calcium carbonate (CaCO3, in percentage) along the pH gradient, in the north of France. Calculations were performed with 2798 relevés for S, 1208 relevés for Al and 333 for CaCO3 (source: EcoPlant & Gégout, 2001).

environments in the NE or the NW to richer environments in the NW where we infer that competitive conditions are weaker. Regional pools of neutrophilous species and competitive conditions (species density) might then explain some of the patterns expressed by Fig. 3. Discussion and conclusions In northern France, the observed low proportion of acid philous vs. basophilous species, coupled with a low proportion of basic

Figure 5 Average numbers of herbaceous species per plot in the NW and NE regions along the pH gradient. Plots were combined into 10 segments in each region.

vs. acidic forest sites, illustrates the ‘calcareous riddle’ described by Ewald (2003) for Central Europe: the fact that there are many acidic forest sites with relatively few adapted species and few neutral to rich forest sites with relatively many adapted species. Our main result, the relative stability of pH niche for most forest species in northern France, is consistent with most previous studies (e.g. Thompson et al., 1993; Hill et al., 2000; Prinzing et al., 2002). Results are however, based on real pH measurements and were derived from two high-quality regional data sets. Even if the use of empirical indicator values, such as Ellenberg’s values, could lead to consistent results (e.g. Diekmann, 1995; Ertsen et al., 1998; Dzwonko, 2001), we share the idea that linking ecological behaviour of forest plant species to measured environmental factors will improve our knowledge of species’ ecology

Table 1 Herbaceous forest species showing a shift in pH optimum or amplitude greater than 0.5 pH units between the NW and NE study areas (see also Fig. 3). OPT is the ecological optimum, CI95 is the 95% confidence interval around the optimum and AMP is the ecological amplitude Radical

OPT NW

CI95 NW

OPT NE

CI95 NE

AMP NW

AMP NE

Shift in both OPT and AMP Convallaria majalis Dryopteris carthusiana Galeopsis tetrahit Oxalis acetosella Teucrium scorodonia

conma drca gate oxac tesc

6.0 5.1 6.2 5.0 4.6

[5.8, 6.5] [3.5, 5.6] [5.8, 7.3] [3.5, 5.6] [4.3, 4.8]

8.1 3.4 5.6 4.5 3.4

— — [5.4, 5.8] [3.9, 4.8] —

3.0 3.2 2.9 2.9 2.1

3.5 2.4 3.5 2.1 2.7

Shift only in OPT Athyrium filix-femina Carex flacca Geranium robertianum Melica uniflora Mercurialis perennis Stachys sylvatica

atfi cafla gero meun mepe stsy

5.7 7.1 6.9 6.6 8.1 6.4

[5.4, 6.2] [6.7, 8.1] [6.6, 7.7] [6.2, 7.8] — [6.1, 6.9]

5.1 8.1 6.3 6.1 7.5 8.1

[4.8, 5.3] — [6.1, 6.7] [5.9, 6.3] [7.1, 8.1] —

3.0 2.4 2.5 2.9 2.0 2.4

2.7 2.0 2.5 2.8 1.8 2.3

Shift only in AMP Carex remota Luzula pilosa Melampyrum pratense Scrophularia nodosa

care lupi mepr scno

5.9 5.5 3.4 6.2

[5.6, 6.5] [5.2, 5.8] — [6.0, 6.6]

5.8 5.4 3.4 6.0

[5.6, 6.2] [5.3, 5.5] — [5.6, 6.5]

2.9 2.7 2.6 2.3

2.4 2.2 1.5 2.9

Global Ecology and Biogeography, 14, 263– 270, © 2005 Blackwell Publishing Ltd

267

C. Coudun and J.-C. Gégout (Schaffers & Sykora, 2000; Wamelink et al., 2002). Diekmann & Lawesson (1999) have characterized the ecological behaviour of eight species in four distinct regions, along a gradient from central Europe to northern Europe (Europe latitudinal gradient) and they showed that some species could reveal geographical shifts in the ecological optimum. This study was broadened by other authors who explored longitudinal gradients in Europe (Hill et al., 1999; Prinzing et al., 2002; Gégout & Krizova, 2003). A further step would be to contrast our findings with existing literature (Ellenberg et al., 1992; Hill et al., 1999; Gégout & Krizova, 2003), to check whether computed pH indicator values in northern France are consistent along a wider Europe longitudinal gradient, from oceanic Great Britain to continental Eastern Europe. Assigning all species a unique pH indicator value valid through all temperate Europe is probably not reasonable, but it would be worth investigating further those aspects of geographical shifts in the ecological behaviour at the European scale by further analyses performed on large databases. We showed that 11 species out of 46 presented a slight shift towards most favourable mineral nutrition environments in the NW (Table 1). However, a technical artefact, pointed out early by Mohler (1983), might be responsible for an artificial shift in the pH optimum and amplitude for some species, by adding variability to the pH niche characterization; we noticed indeed that those species that have an optimum at an extreme of the pH gradient in one region often presented a shift in the optimum (e.g. Dryopteris carthusiana, Teucrium scorodonia at the lower extreme of the pH gradient and Mercurialis perennis, Convallaria majalis, Stachys sylvatica and Carex flacca at the higher extreme of the pH gradient). Results reveal that the shift in optimum for these species is coherent, for five species out of six, with the assumption of a shift towards favourable resource conditions in the NW. This artefact is likely to affect the shift extent but is not likely to bias the main results. Models fitted for the same species, but in different areas are difficult to compare (Guisan et al., 2002) because they basically model the realized niche and implicitly take into account different abiotic contexts and biotic interactions that are likely to change between regions (Guisan & Zimmermann, 2000). In our study, however, we tried to neutralize the effect of environmental factors other than pH (mineral nutrition) to avoid investigating ecological shifts of species’ behaviour in too varied an array of abiotic contexts (above). The fact that the relations between pH and other nutrition factors such as base saturation rate or C/N ratio are stable between the NW and the NE allows us to reject the hypothesis of a hidden factor to explain the slight shifts observed for some species. The assumption of ecotypic differentiation for most species between the NW and the NE was not realistic because it could not be responsible for the systematic shifts observed in our results. The few observed shifts were thus argued to be due to different biotic contexts in our two regions. Those trends for greater pH optimum and/or amplitude values were possibly due to competitive release conditions in NW neutral to rich environments, and we suggest that conversely, plant species seem to be submitted to harsher competition conditions in NE forest sites whose 268

pH values lie between six and seven. On average, there are three more species per relevé in the NE than in the NW on neutral to basic soils (Fig. 5), and we also observed that the total herbaceous cover in the NE exceeds 80% from pH value of six and up, and is greater than the total herbaceous cover in the NW forests, by at least 5% cover, all along the pH gradient. A possible explanation for this could be a higher density of forests in the NE than in the NW and thus facilitated plant mobility and colonization (see Fig. 1). Our study dealt with the totality of frequent herbaceous species in northern France’s forests, and generated results that could not have been found by studying only a couple of species. The influence of competition on the ecological response of species has rarely been shown with real data and in this study has been indirectly inferred: further investigations should be promoted in order to test these ecological theories. ACKNOWLEDGEMENTS The authors wish to thank Christian Piedallu for his appreciated technical support, Anna Ramshaw for her careful correction of the English language, Jean-Christophe Hervé and Jean-Luc Dupouey for stimulating discussions, as well as Ole R. Vetaas and another anonymous reviewer for their appropriate comments on earlier drafts of the manuscript. This study was financed through grants to Christophe Coudun by the French National Forest Office (ONF) and the Lorraine Regional Council (CR Lorraine). EcoPlant is a phytoecological database financed by the French Institute of Agricultural, Forest and Environmental Engineering (ENGREF), the French Ministry of Agriculture (DERF) and the French Agency for Environment and Energy Management (ADEME). REFERENCES Akaike, H. (1973) Information theory as an extension of the maximum likelihood principle. Second symposium on information theory (ed. by B.N. Petrov and F. Csaki), pp. 267–281. Akademiai Kiai, Budapest. Austin, M.P. (1980) Searching for a model for use in vegetation analysis. Vegetatio, 42, 11–21. Austin, M.P. (1987) Models for the analysis of species’ response to environmental gradients. Vegetatio, 69, 35 –45. Austin, M.P. (2002) Spatial prediction of species distribution: an interface between ecological theory and statistical modelling. Ecological Modelling, 157, 101–118. Austin, M.P., Cunningham, R.B. & Fleming, P.M. (1984) New approaches to direct gradient analysis using environmental scalars and statistical curve-fitting procedures. Vegetatio, 55, 11–27. Austin, M.P. & Nicholls, A.O. (1997) To fix or not to fix the species limits, that is the ecological question: response to Jari Oksanen. Journal of Vegetation Science, 8, 743–748. Badeau, V. (1998) Caractérisation écologique du Réseau européen de suivi des dommages forestiers. Bilan des opérations de terrain et premier résultats. Les cahiers du Département

Global Ecology and Biogeography, 14, 263–270, © 2005 Blackwell Publishing Ltd

Ecological behaviour of herbaceous forest species along a pH gradient de la Santé des Forêts, 5. Ministère de l’Agriculture et de la Pêche, Direction de l’Espace Rural et de la Forêt, Paris. Bessemoulin, J. (1989) Atlas climatique de la France, édition réduite. Direction de la Météorologie Nationale, Paris. ter Braak, C.J.F. & Gremmen, N.J.M. (1987) Ecological amplitudes of plant species and the internal consistency of Ellenberg’s indicator values for moisture. Vegetatio, 69, 79 –87. ter Braak, C.J.F. & Looman, C.W.N. (1986) Weighted averaging, logistic regression and the Gaussian response model. Vegetatio, 65, 3–11. Diekmann, M. (1995) Use and improvement of Ellenberg’s indicator values in deciduous forests of the Boreo-nemoral zone in Sweden. Ecography, 18, 178 –189. Diekmann, M. (1996) Ecological behaviour of deciduous hardwood trees in Boreo-nemoral Sweden in relation to light and soil conditions. Forest Ecology and Management, 86, 1–14. Diekmann, M. (2003) Species indicator values as an important tool in applied plant ecology: a review. Basic and Applied Ecology, 4, 493 –506. Diekmann, M. & Lawesson, J.E. (1999) Shifts in ecological behaviour of herbaceous forest species along a transect from northern central to north Europe. Folia Geobotanica, 34, 127– 141. Duchaufour, P. (1989) Pédologie et groupes écologiques. I. Rôle du type d’humus et du pH. Bulletin d’Ecologie, 20, 1– 6. Dzwonko, Z. (2001) Assessment of light and soil conditions in ancient and recent woodlands by Ellenberg indicator values. Journal of Applied Ecology, 38, 942–951. Ellenberg, H., Weber, H.E., Düll, R., Wirth, V., Werner, W. & Paulißen, D. (1992) Zeigerwerte von Pflanzen in Mitteleuropa. Scripta Geobotanica, 18, 1–248. Ertsen, A.C.D., Alkemade, J.R.M. & Wassen, M.J. (1998) Calibrating Ellenberg indicator values for moisture, acidity, nutrient availability and salinity in the Netherlands. Plant Ecology, 135, 113 –124. Espiau, P. & Peyronel, A. (1977) Acidité d’échange dans les sols. Application à une séquence altitudinale des sols du massif du Mont Aigoual. Science du Sology, 4, 25 – 44. Ewald, J. (2003) The calcareous riddle: why are there so many calciphilous species in the Central European flora? Folia Geobotanica, 38, 357–366. Falkengren-Grerup, U., Brunet, J., Quist, M.E. & Tyler, G. (1995) Is the Ca: Al ratio superior to pH, Ca or Al concentrations of soils in accounting for the distribution of plants in deciduous forests? Plant and Soil, 177, 21–31. Gégout, J.-C. (2001) Création d’une base de données phytoécologiques pour déterminer l’autécologie des espèces de la flore forestière de France. Revue Forestière Française, 53, 397– 403. Gégout, J.-C. & Krizova, E. (2003) Comparison of indicator values of forest understory plant species in Western Carpathians (Slovakia) and Vosges Mountains (France). Forest Ecology and Management, 182, 1–11. Gégout, J.-C. & Pierrat, J.-C. (1998) L’autécologie des espèces végétales: une approche par régression non paramétrique. Ecologie, 29, 473 –482.

Gough, L., Shaver, G.R., Carroll, J., Royer, D.L. & Laundre, J.A. (2000) Vascular plant species richness in Alaskan arctic tundra: the importance of soil pH. Journal of Ecology, 88, 54 –66. Guisan, A., Edwards, T.C. Jr & Hastie, T.J. (2002) Generalized linear and generalized additive models in studies of species distributions: setting the scene. Ecological Modelling, 157, 89–100. Guisan, A. & Zimmermann, N.E. (2000) Predictive habitat distribution models in ecology. Ecological Modelling, 135, 147–186. Hill, M.O., Mountford, J.O., Roy, D.B. & Bunce, R.G.H. (1999) Ellenberg’s indicator values for British plants. Technical annex to volume 2 of the ECOFACT research report series, Centre for Ecology and Hydrology, Monks Wood, UK. Hill, M.O., Roy, D.B., Mountford, J.O. & Bunce, R.G.H. (2000) Extending Ellenberg’s indicator values to a new area: an algorithmic approach. Journal of Applied Ecology, 37, 3–15. Lawesson, J.E. & Mark, S. (2000) pH and Ellenberg reaction values for Danish forest plants. Proceedings of the 41st IAVS Symposium: Vegetation science in retrospective and perspective (ed. by P.S. White, L. Mucina and J. Leps), 151–153. Opulus Press, Uppsala, Sweden. Lawesson, J.E. & Oksanen, J. (2002) Niche characteristics of Danish woody species as derived from coenoclines. Journal of Vegetation Science, 13, 279–290. Le Tacon, F. & Timbal, J. (1972) A propos des conditions écologiques des hêtraies dans le nord-est et le nord-ouest de la France. Revue Forestière Française, 24, 187–200. Lehmann, A., Overton, J.McC. & Austin, M.P. (2002) Regression models for spatial prediction: their role for biodiversity and conservation. Biodiversity and Conservation, 11, 2085–2092. MathSoft, Inc. (1999) S-Plus 2000: programmer’s guide. MathSoft, Inc, Seattle. McCullagh, P. & Nelder, J.A. (1997) Generalized linear models, 2nd edn. Chapman & Hall, London. Mohler, C.L. (1983) Effect of sampling pattern on estimation of species distributions along gradients. Vegetatio, 54, 97–102. Odland, A., Birks, H.J.B. & Line, J.M. (1995) Ecological optima and tolerances of Thelypteris limbosperma, Athyrium distentifolium, and Matteuccia struthiopteris along environmental gradients in Western Norway. Vegetatio, 120, 115–129. Oksanen, J. (1997) Why the beta-function cannot be used to estimate skewness of species responses. Journal of Vegetation Science, 8, 147–152. Oksanen, J., Läärä, E., Tolonen, K. & Warner, B.G. (2001) Confidence intervals for the optimum in the Gaussian response function. Ecology, 82, 1191–1197. Parrish, J.A.D. & Bazzaz, F.A. (1985) Ontogenetic niche shifts in old-field annuals. Ecology, 66, 1296–1302. Prinzing, A., Durka, W., Klotz, S. & Brandl, R. (2002) Geographic variability of ecological niches of plant species: are competition and stress relevant? Ecography, 25, 721–729. Rameau, J.-C., Mansion, D., Dumé, G., Lecointe, A., Timbal, J., Dupont, P. & Keller, R. (1993) Flore forestière française. Guide écologique illustré, 2, Montagnes. Institut pour le Développement Forestier, Paris. Rameau, J.-C., Mansion, D., Dumé, G., Timbal, J., Lecointe, A., Dupont, P. & Keller, R. (1989) Flore forestière française. Guide

Global Ecology and Biogeography, 14, 263– 270, © 2005 Blackwell Publishing Ltd

269

C. Coudun and J.-C. Gégout écologique illustré, 1, Plaines et Collines. Institut pour le Développement Forestier, Paris. Schaffers, A.P. & Sykora, K.V. (2000) Reliability of Ellenberg indicator values for moisture, nitrogen and soil reaction: a comparison with field measurements. Journal of Vegetation Science, 11, 225–244. Scott, J.M., Heglund, P.J., Samson, F., Haufler, J., Morrison, M., Raphael, M. & Wall, B. (2002) Predicting species occurrences: issues of accuracy and scale. Island Press, Covelo. Thompson, K., Hodgson, J.G., Grime, J.P., Rorison, I.H., Band, S.R. & Spencer, R.E. (1993) Ellenberg numbers revisited. Phytocoenologia, 23, 277–289. Turesson, G. (1922) The species and the variety as ecological units. Hereditas, 3, 100 –113. Tutin, T.G., Heywood, V.H., Burges, N.A., Valentine, D.H., Walters, S.M. & Webb, D.A. (2001) Flora Europaea, vol. 1–5. Cambridge University Press, Cambridge. Tyler, G. (2003) Some ecophysiological and historical approaches to species richness and calcicole/calcifuge behaviour — contribution to a debate. Folia Geobotanica, 38, 419 – 428. Vetaas, O.R. (2000) Comparing species temperature response curves: population density versus second-hand data. Journal of Vegetation Science, 11, 659– 666.

270

Walter, H. & Walter, E. (1953) Einige allgemeine Ergebnisse unserer Forschungsreise nach Südwestafrika 1952/1953: das Gesetz der relativen Standortskonstanz; das Wesen der Pflanzengemeinschaften. Berichte der Deutshen Botanischen Gessellschaft, 66, 227–235. Wamelink, G.W.W., Joosten, V., Van Dobben, H.F. & Berendse, F. (2002) Validity of Ellenberg indicator values judged from physico-chemical field measurements. Journal of Vegetation Science, 13, 269–278. BIOSKETCHES Christophe Coudun, Environmental Engineer and PhD student in Forest Ecology, is interested in modelling species–environment relationships. He is mainly investigating whether in France, forest plant species reveal geographical shifts in their ecological response to edaphic and climatic factors. Jean-Claude Gégout is an ecologist whose research interests include forest plant ecology and sociology, as well as geographical information systems. He is the main author of the French forest phytoecological database EcoPlant.

Global Ecology and Biogeography, 14, 263–270, © 2005 Blackwell Publishing Ltd