SEARCH

SEARCH BY CITATION

Keywords:

  • dyke;
  • flooding;
  • floodplain;
  • grassland species;
  • levee;
  • logistic regression;
  • regulated river

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information
  • 1
    One of the main targets of river regulation with dams and dykes is the stabilization of highly fluctuating water tables. While there is information about the overall impact of such regulation measures on plant species composition and richness, far less is known about specific species’ response patterns to reduced water level fluctuations.
  • 2
    The response of 30 common grassland species to soil moisture and water level fluctuations was assessed. Floristic data were collected from the floodplain of the Elbe River, Germany, from 182 plots, 99 within the recent floodplain and 83 in an older floodplain, separated from one another by dykes. Hydrological data were collected weekly over 2 years at 37 water level wells. Using logistic regression, the patterns of species’ responses to hydrological regulation were predicted.
  • 3
    The majority of species responded significantly to water level fluctuations. Species of high elevation habitats occurred at lower elevations where water level fluctuations were reduced, indicating increasing drought at high elevation habitats. However, species that occurred in floodplain depressions tended to shift from lower to higher elevations to avoid permanent inundation.
  • 4
    Almost half of the species showed a significant preference for either highly fluctuating water tables, characteristic of the recent floodplain, or for stable water tables, characteristic of the older floodplain. The probability of their occurrence was either reduced or increased along a gradient of reduced fluctuations. These species’ responses could be partly explained by altered flooding regimes, although other factors, such as disturbance and dispersal processes, were also involved.
  • 5
    Synthesis and applications. This study demonstrates that reduced water level fluctuations caused by the construction of dams and dykes lead to substantial changes in the spatial distribution of floodplain plant species and in species composition. The methodology reported here allows accurate prediction of shifts in floodplain vegetation in response to human-induced alterations in floodplain landscapes. This can be used as a tool to assess river regulation measures and for floodplain restoration purposes, such as dyke relocations.

Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

Regulation and fragmentation of river–floodplain systems by dykes and dams, to control flooding, facilitate navigation and make use of hydroelectric power, have altered the flow patterns of nearly all of the major river–floodplain systems in Europe and throughout the northern Hemisphere (Petts 1984, 1989; Dynesius & Nilsson 1994; Nilsson & Berggren 2000; Tockner & Stanford 2002). River regulation measures have strong impacts on riparian vegetation. River damming severely alters species richness and composition, because of the change in hydrological regimes and disruption of the longitudinal flow (Auble, Friedman & Scott 1994; Nilsson & Jansson 1995; Nilsson, Jansson & Zinko 1997; Jansson, Nilsson & Renöfält 2000; Merritt & Cooper 2000). The loss and degradation of natural floodplains caused by dykes (also called levees) are considerable. These structures are usually several metres high and run parallel to the course of a river, preventing flooding of the adjoining countryside as well as restricting the river to its main course. The renewal of fluvial landforms, an important process in natural floodplains, is constrained. Nearly all large rivers and their floodplains in central Europe and the USA are affected by dykes, with an estimated 40 000 km of dykes in the USA alone (Johnston Associates 1989). In central Europe, the floodplain landscapes of the Upper and Lower Rhine River, the Lower Danube River and the Middle Elbe River, for example, have lost more than three-quarters of their natural inundation area as a result of dykes (Dister 1991; Schneider 1991; Dahl & Flade 1994). The ecological impacts of dykes have largely been ignored, although recent studies have demonstrated that separating floodplains from their river channels results in a profound change in floodplain structure, process and species composition (Marston et al. 1995; Girel, Garguet-Duport & Pautou 1997; Trémolières et al. 1998; Deiller, Walter & Trémolières 2001; Gergel, Dixon & Turner 2002; Leyer 2004).

The main effect of a dam is to disrupt longitudinal river flow, whereas dykes disrupt the hydrological connectivity between river and floodplain. Both structures alter natural water level fluctuation patterns: the seasonal flow variability and peak flows are considerably reduced. Therefore, river regulation measures are invariably accompanied by declining water level fluctuations. The alternating wet–dry cycle of periodically flooded areas is of great importance for riparian and lake vegetation (Crawford 1996; Davis et al. 1996; Gowing & Spoor 1998; Hill, Keddy & Wisheu 1998). Few studies, however, have analysed quantitative data on water level fluctuations in relation to riparian vegetation patterns, and there is an almost complete lack of knowledge on specific plant species’ responses to reduced water level fluctuations. A pioneering study by Johansson & Nilsson (2002) compared the growth and survival of transplanted individuals of four riparian plant species on free-flowing and regulated riverbanks in Sweden. They found a clear reduction in the performance of plants growing on banks of regulated rivers because of changes in the water level regime.

This study investigated the response of floodplain species to reduced water level fluctuations on a gradient from highly fluctuating to stable water tables. Highly fluctuating regimes are typical of natural floodplains, being characterized by a considerable flooding depth in the high water season and a dry period with a strong decrease in groundwater in the low water season. Stable conditions are typical of floodplains that have been isolated from the river flow by dykes. Hydrological parameters, including average groundwater level, flooding duration and flooding depth, as well as disturbance and dispersal processes, will influence the response patterns of riparian vegetation. The aim of this study was to predict shifts in floodplain vegetation in response to human-caused alterations in floodplain landscapes and to provide information for an assessment of river regulation and floodplain restoration measures.

Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

study area

With a length of about 1096 km and a catchment area of approximately 148 000 km2, the Elbe River in Germany is one of the large rivers in central Europe. It originates in the Giant Mountains and passes through the Czech Republic (Upper Elbe) and Germany before discharging in the North Sea. Glacial processes during the Ice Age shaped the river valley into a large lowland section of about 640 river km.

The study was carried out in a part of the lowland river valley that belongs to the Middle Elbe River, half way between Hamburg and Berlin (380–470 Elbe km, the most southerly point at 52°24′N, 12°00′W, the most northerly point at 53°03′N, 11°30′W). The flow regime can be characterized as rain-fed/snow-fed, with the highest water levels usually attained in March–April and the lowest between August and October. The mean annual river discharge of the Elbe in this area is about 600 m3 s−1. The width of the entire floodplain ranges between 3 and 19 km and is bordered by a belt of glacial sands and moraines. The height above sea level varies from about 40 m in the south to 15 m in the north.

The Upper Elbe in the Czech Republic is characterized by more than 20 weirs and navigation locks, and one impoundment at Geesthacht (Germany), 110 km downstream of the study area, divides the lowland river from the tidal river section. The Middle Elbe River has no dams interrupting the flow of water, so the floodplain is still strongly influenced by seasonal water level fluctuations. The first dykes were built in the 12th century and flood protection activities were intensified during the 18th century, resulting in a continuous dyke line in the early 1900s. At present the dyke line (often more than 9 m high) prevents flooding in more than 75% of the floodplain in the study area (Dahl & Flade 1994). Therefore the entire floodplain can be divided into a recent floodplain, which is still flooded by river water, and an older floodplain, which was flooded before the construction of dykes.

The recent floodplain is characterized by a high variation in relief and in higher average elevations than the older floodplain. The latter can be inundated in its lower parts as well, because of increasing groundwater in times of high water. The inundation can be long-lasting, but it is not as deep as occurs in the recent floodplain. Although water level fluctuations have generally decreased in the older floodplain, some areas near the dykes show considerable groundwater fluctuations (amplitudes up to 2·5 m recorded during this study), while the fluctuations at the floodplain margin are restricted to an amplitude of 1 m. The recent floodplain is characterized by higher fluctuating water tables (up to 5 m). Grain sizes in the alluvial soil sediments are highly variable and correlate with elevation. Sandy sediments are mainly deposited at higher elevations and the amount of silt and clay increases with declining elevation.

A long period of human colonization (> 3000 years) has resulted in a transformation of the original floodplain into large areas of cut and grazed grasslands as well as arable fields. Only small patches of the natural floodplain forests remain in the study area. The grassland vegetation is well differentiated along an elevation gradient of increasing soil moisture and flooding frequency. Furthermore, vegetation patterns differ between the recent and older floodplains. The higher elevations of the floodplains are occupied by a xeric grassland community (DianthoArmerietum elongatae, phytosociological class Koelerio–Corynephoretea), which is replaced by communities of the class Molinio–Arrhenatheretea at intermediate elevations. While in the recent floodplain Alopecurus pratensis-dominated meadows prevail, in the older floodplain communities of the alliance Arrhenatherion, Cynosurion and Cnidion are typical. The areas of low elevation in the recent floodplain and the older floodplain near the dykes are covered by flood meadows of the class Agrostietea stoloniferae, typically characterized by Alopecurus geniculatus, Agrostis stolonifera, Eleocharis uniglumis and Carex vulpina. These meadows are displaced by Caltha palustris–swamp meadows in habitats of stable water tables. These meadows mainly occur at the floodplain margin, which belongs to the older floodplain (Leyer 2002).

Species nomenclature follows that of Wisskirchen & Haeupler 1998).

hydrological monitoring and analysis

In October 1996, 37 water level wells were installed in grassland areas throughout the recent and older floodplain along a 90-km stretch of the Middle Elbe River floodplain to generate precise hydrological parameters. Different elevations were sampled, as well as different distances to the Elbe River and the dyke line, in order to collect data along a vertical and horizontal cross-section of the floodplain. The wells were examined approximately weekly between November 1996 and February 1999. Each well consisted of a 2-m length of 2·5-cm diameter PVC pipe inserted with its whole length into the ground. A free flow of water from the inside to the outside and vice versa was ensured by several slits along the lower 1·7 m. For the purpose of surface water level measurements and to mark the well within the area, a second 2-m pipe at a distance of 50 cm was installed, which was inserted half-way into the ground. A 1-m scale on the upper half was used to determine the water level during floods, with binoculars or by boat. Two or more wells were installed at different elevations in areas characterized by a high variation in relief. Therefore, measurements were ensured both at flooding times (when the lower pipe was underwater) and at times with low groundwater levels (when the groundwater was below the upper well).

The weekly data were transformed to a continuous water level line for the years 1997 and 1998 using regression. The explanatory variable was the water level of the Elbe River (daily values), based on records from the official gauging stations Wittenberge, Neuwerben and Tangermünde. The delay of 1–20 days in the response of ground water levels to the flow pattern of the Elbe River was compensated for by a correcting time shift before performing the calculations. In addition, the groundwater data showed a smoother line than the river water fluctuation line, so the moving average technique was used to take this difference into account (Crawley 2003). The water level line for each of the 182 plots in which the vegetation was sampled (see below), was derived from that of the nearest well, including the altitude of the plot in relation to the well. Four hydrological parameters were calculated for each plot using water level data from the water level line for 1997 and 1998. (i) Average groundwater level (cm): arithmetic mean, related to the soil surface of every vegetation plot studied. The values are usually negative because the average groundwater level lies under the soil surface. Very low values indicate plots of high elevation, while values around zero represent floodplain depressions. (ii) Size of water level fluctuations (cm): the standard deviation of the water level line; high values represent highly fluctuating water tables, while stable water tables are characterized by low values (Fig. 1). (iii) Duration of flooding (days year−1). (iv) Duration of water levels higher than 50 cm above the soil surface (days year−1): 50 cm as a rough value for a flooding depth, at which average plants are completely under water.

image

Figure 1. Representative example of water level variation in the recent and older floodplain (centred to zero mean) and the corresponding stage data of the Elbe River, Germany.

Download figure to PowerPoint

vegetation sampling

In May and June from 1996 to 1998, grassland vegetation was sampled in 182 plots, each 12 m2 in size, distributed over the entire floodplain near the water level wells. In areas with a high variation in relief, up to eight plots were established per well, while in sites with little variation in relief the number of plots per well was 1–2. The recent floodplain encompassed 99 plots, while the older floodplain encompassed 83 plots. The vegetation of both floodplain types was sampled every year. Grasslands created by artificial seeding or those that were intensively fertilized were excluded. Plant species cover was estimated using the Londo scale (Londo 1976), which partitions the cover range into 13 classes. The plot position was marked with a permanent magnet, surveyed by compass and recorded in a sketch. Using a magnetic locator, the plots could be retrieved and levelled.

analysis

Logistic regression

Thirty of the 50 most frequent species of the data set were randomly selected for analysis of the effects of reduced water level fluctuations on grassland vegetation. Preference models for these species were developed by logistic regression, describing the probability of occurrence of a species as a function of one or more environmental variables (ter Braak & Looman 1995). Prior to the calculations, the species abundance data were transformed into presence–absence data. The data sets were split into two halves, a training subset and a testing subset. The first was used to develop the prediction model, and the second to estimate the robustness of prediction (cross-validation; Huberty 1994; Fielding & Bell 1997). Average groundwater level (AGL) and size of water level fluctuations (WLF) were used as independent environmental variables, which, in combination, strongly govern grassland species composition at the Elbe River (Leyer 2004). As a result of either unimodal or sigmoid species–environmental relationships, linear models with or without a quadratic term were selected. Further tests were conducted to see if an interaction term (AGL × WLF) needed to be included in the model. Preference models were evaluated by the explained deviance. The significance of a chosen model in comparison with the null model or a different model (unimodal against sigmoid; AGL + WLF against AGL; AGL + WLF + interaction term against AWL + WLF) was tested by a chi-square test (P < 0·05, Crawley 2003).

Model discrimination and validation

The calculated probabilities of occurrence were used to estimate the prediction power of the models by performing two techniques. The first one was based on a classification matrix, which compares the predicted values of a model with the observed data from which the model was derived. This method requires a threshold value of probability (pcrit). If the calculated value p is less than pcrit, the habitat is defined as not occupied, and vice versa. Using the numbers of correct and false predicted occurrences and absences, performance criteria can be described, such as sensitivity (proportion of correctly predicted occurrences) and specificity (proportion of correctly predicted absences) and the correct classification rate (Fielding & Bell 1997). Here a threshold value popt was selected, which maximizes the proportion of correct classifications (Zweig & Campbell 1993). Furthermore, another threshold value, pfair, was used, at which sensitivity and specificity are the same (Schröder & Richter 1999). The G-test (log-likelihood ratio test; Sokal & Rohlf 1995) was used to test the significance of the predicted and observed data in the classification matrix based on pfair (Thomas & Bovee 1993; Schröder 2000). It tested the null hypothesis that suitable habitats (prognosis: presence) and unsuitable habitats (prognosis: absence) are occupied proportionally.

One of the problems with the threshold-dependent measures is their failure to use all of the information provided by the classifier (Fielding & Bell 1997) in addition to a certain arbitrariness by choosing optimal thresholds (Altman et al. 1994). Therefore, a second technique, the receiver operating characteristic (ROC) technique, was performed as a threshold-independent method to estimate the predictive power of the models (Zweig & Campbell 1993). A ROC curve is obtained by plotting all sensitivity values on the y-axis against their equivalent (1 − specificity) values for all available thresholds on the x-axis. The area under the ROC function (AUC) is an important index, because it provides a single measure of overall accuracy independent of a particular threshold (Fielding & Bell 1997). An AUC value of 0·5 indicates a random model without any power of prediction; values between 0·7 and 0·8 represent acceptable models; values between 0·8 and 0·9 represent excellent models; and higher values represent outstanding models (Hosmer & Lemeshow 2000). The AUC value is presented here with the corresponding confidence interval (generated by bootstrapping; Zweig & Robertson 1982) as well as the significance level. It was calculated by (AUC − AUCcrit)/SEAUC .The SE was calculated according to Hanley & McNeil (1982). The algorithm tested against the null hypothesis that the AUC values do not differ significantly from AUCcrit= 0·5, which represents an area under the ROC curve of a random model (Zweig & Campbell 1993; Schröder 2000).

Further analysis

The optimum (u) of unimodal species response curves was calculated using the regression parameters of each model (ter Braak & Looman 1986). In the case of sigmoid responses instead of the optimum, either the lowest or the highest value of the considered gradient was taken, dependent on the maximum probability of occurrence. This value is termed optimum as well. Contingency tables were used to establish whether the occurrence of species was independent of the floodplain type (recent and older). The significance was tested by applying the G-test. Hydrological differences between the recent and older floodplain were determined by the Mann–Whitney U-test. AUC values were calculated with the software ROC_AUC (Schröder 2004), all other analysis were made with R (R-Development Core Team 2004).

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

Hydrological variables differed between the recent and older floodplains (Table 1). The recent floodplain varied widely in average groundwater level and had a lower median value than the older floodplain (U-test, P < 0·0001). Furthermore, the recent floodplain was characterized by significantly higher fluctuating water tables (U-test, P < 0·0001) and a greater number of water levels higher than 50 cm above the soil surface (U-test, P < 0·0001). There was no difference in the flooding duration of the floodplain types.

Table 1.  Comparison of hydrological variables between the floodplain types ‘recent’ and ‘older’ calculated over the time period 1997 and 1998
Hydrological variableAverage groundwater level (cm)Size of water level fluctuations (cm)Duration of flooding (days year−1)Water levels > 50 cm (days year−1)
RecentOlderRecentOlderRecentOlderRecentOlder
Minimum−298−14950·716·4  0  0  0 0
25 percentile value−200 −8760·829·8  6·5  0  2 0
Median−134 −6370·438·3 17 30  9 0
75 percentile value −75 −3977·140·0 49 71 17 0
Maximum    15  −478·564·720418711421

Within the 182 plots, 234 plant species were identified. Species richness did not exceed 49 species 12-m−2 plot, with a median value of 22. The frequency in occurrence of the 30 species that were selected for analysis ranged between 20% and 68%, with a median value of 60%.

Sigmoid or unimodal models of logistic regression were developed for each species in relation to the average groundwater level (AGL). Most of the species tested were strongly related to this parameter (Table 2). Only Rumex acetosa showed no significant relation; however, in combination with the water level fluctuation gradient (WLF), multiple logistic regression detected the independent importance of AGL, even for Rumex acetosa. For all species, the explained deviance varied widely and accounted, on average (median), for about 20% of the total deviance. The explained deviance significantly increased if the water level fluctuation parameter (in its linear form) was also included in the model. Only five species (Carex praecox, Glechoma hederacea, Vicia cracca, Poa trivialis and Rumex thyrsiflorus) were not affected by the water level fluctuation gradient, even in those cases where it was only added in the regression model. The occurrences of these species seemed to be independent of water level fluctuation changes, and were therefore not considered in further analysis. The explained deviance of the combination AGL + WLF (with or without an interaction term consisting of both parameters) for all other species ranged from 14% to 59% of the total deviance and had a median value of about 33%.

Table 2.  Results of logistic regression analyses for species in relation to AGL and the combination with WLF (AGL + WLF). AUC values and classification rates based on popt and pfair are displayed as criteria for the predictive power of the models. ***P < 0·001; **P < 0·01; *P < 0·05
 Training data setTesting data set
Total devianceAGL explained deviance (%)AGL + WLF explained deviance (%)AUC (CI)Correct (%, popt)Correct (%, pfair)
Achillea millefolium102·927·0***49·2***0·91 (0·83, 0·96)***91·280·2***
Agrostis capillaris116·839·4***48·6**0·94 (0·89, 0·98)***86·885·7***
Agrostis stolonifera112·316·8***20·0*0·73 (0·64, 0·83)**73·672·4***
Alopecurus geniculatus 93·332·8***54·0***0·90 (0·85, 0·96)***83·582·4***
Alopecurus pratensis108·921·0***31·4*0·85 (0·76, 0·93)***80·279·1***
Cardamine pratensis105·013·3***23·1**0·79 (0·68, 0·88)***75·872·5***
Carex acuta105·033·2***36·7*0·85 (0·75, 0·92)***80·279·1**
Carex vulpina 87·725·5***33·1**0·84 (0·72, 0·95)***84·678·0***
Cerastium holosteoides118·010·1**14·0*0·84 (0·76, 0·91)***78·073·6***
Cnidium dubium100·710·3**17·5*0·84 (0·74, 0·93)***84·682·4***
Deschampsia cespitosa102·916·2***32·9***0·85 (0·76, 0·92)***85·779·1***
Elymus repens126·133·1***38·2*0·92 (0·85, 0·97)***86·986·9***
Festuca pratensis107·0 5·6*19·4***0·83 (0·72, 0·92)***82·478·0***
Galium verum agg. 93·239·3***49·5**0·93 (0·86, 0·97)***87·986·8***
Holcus lanatus102·919·7***35·6***0·81 (0·65, 0·91)***90·170·3**
Persicaria amphibia105·017·8***26·3**0·85 (0·77, 0·93)***82·472·5***
Phalaris arundinacea121·320·8***25·3*0·79 (0·69, 0·88)***75·875·8***
Plantago intermedia 78·1 6·2*15·0**0·78 (0·67, 0·89)***84·671·4**
Poa palustris116·815·7**29·1***0·85 (0·75, 0·90)***76·974·7***
Poa pratensis agg.125·346·4***59·0***0·88 (0·79, 0·95)***84·684·6***
Potentilla anserina112·329·8***33·2*0·89 (0·86, 0·91)***84·683·5***
Ranunculus repens122·220·5***45·8***0·82 (0·72, 0·90)***79·173·6***
Rorippa sylvestris 92·311·3**35·8***0·85 (0·75, 0·92)***81·378·0***
Rumex acetosa 95·9– NS29·5***0·89 (0·87, 0·95)***81·178·9***
Stellaria palustris116·823·9***31·6*0·88 (0·80, 0·95)***83·577·4***

Results of the ROC technique (AUC values), as well as the number of correct classifications based on two cut-off values (popt and pfair), were generated both for the training and testing data to discriminate and validate the models. On the whole, the AUC values of the training data were slightly better because they were optimally adjusted. This was also true for the classification results. In further analyses and discussion, only the results of the testing data set were considered, which represented a more realistic estimate of prediction accuracy.

The models consisting of AGL (sigmoid or unimodal), WLF (sigmoid) and partly an interaction term yielded highly significant results (Table 2). The AUC values (≥ 0·73) reflected the high predictive power of each model. The number of correct classifications supported these results.

Two groups of species’ response patterns along a gradient of water level fluctuations could be distinguished. The first group was characterized by the fact that the species’ optima towards AGW changed along a gradient of WLF (Fig. 2). Most species belonging to this group had a preference for a lower AGL where the water tables fluctuated widely, in comparison with the optima under stable water tables. While the first group had similar occurrence probabilities over the whole range of water level fluctuations, the second group responded to reduced fluctuations with an increase or decrease in the probability of occurrence (Fig. 3). Some species from the second group had a tendency towards different AGL optima (Deschampsia cespitosa, Stellaria palustris and Carex vulpina). In the strict sense, these species belonged to both groups of response patterns.

image

Figure 2. Species’ probability of occurrence in relation to the size of WLF and the AGL. The latter correspond to an elevation gradient. First group of species’ response patterns: change of optimum AGL along a gradient of fluctuations.

Download figure to PowerPoint

image

Figure 3. Species’ probability of occurrence in relation to the size of WLF and the AGL. Second group of species’ response patterns: declining or increasing probability of occurrence along the gradient of reduced WLF.

Download figure to PowerPoint

To examine species’ responses to the fluctuation gradient at a finer scale, the data set was partitioned into two subsets: the recent and the older floodplains. As both floodplain types were well differentiated with regard to the fluctuation gradient (Table 1), the two groups of species’ response patterns should also be reflected at this scale.

The majority of the species was significantly related to the AGL gradient (see the Appendix). Species’ response curves strongly differed between the recent and the older floodplains. Species belonging to the first group preferred different AGL values in the recent floodplain compared with the older floodplain (e.g. Poa palustris and Galium verum agg.; Fig. 4a,b). Species from the second group either had a quite high probability of occurrence within the older floodplain, while they rarely occurred in the recent floodplain (e.g. Deschampsia cespitosa and Rumex acetosa; Fig. 4c) or vice versa (e.g. Rorippa sylvestris and Plantago intermedia). Some species were not significantly related to AGL in the recent floodplain because of a complete or almost complete absence (e.g. Cardamine pratensis, Holcus lanatus and Potentilla anserina). These species belonged to the second group as well.

image

Figure 4. (a–c) Examples of species’ response curves in relation to AGL for the recent (dotted line) and older (solid line) floodplains.

Download figure to PowerPoint

The results confirmed the patterns that were detected by analysing the whole data set. Generally, the results of the contingency tables, which were used to assess whether the occurrence of each species was independent of the floodplain type, conformed to the described patterns as well. Eight species revealed significant associations with the older floodplain (Cardamine pratensis, Carex acuta, Deschampsia cespitosa, Festuca pratensis, Holcus lanatus, Potentilla anserina, Rumex acetosa and Stellaria palustris) and three with the recent floodplain (Elymus repens, Plantago intermedia and Rorippa sylvestris), while the remainder did not exhibit a relationship with a particular floodplain type at all. Species with a declining probability of occurrence along a gradient of declining water level fluctuations were significantly more frequent in the recent floodplain and vice versa. Only in the case of Alopecurus geniculatus and Carex vulpina did the G-test not recognize a significant preference for floodplain type, while the response patterns detected by logistic regression displayed the opposite (Fig. 3). Carex acuta represented a converse case.

The optimum (u) of the AGL gradient was determined for each species’ response curve of the recent and older floodplains. Species without a model for each floodplain type were not considered. The relation between the difference (urecent − uolder) and the optimum of the recent floodplain, urecent, showed the direction of species’ responses (Fig. 5).

image

Figure 5. Differences in species’ optima between recent and older floodplains in relation to the species’ optima of the recent floodplain.

Download figure to PowerPoint

Species preferring low average groundwater levels in the recent floodplain (indicating high elevation habitats) selected habitats with much higher average groundwater levels in the older floodplain (indicating a preference for lower elevation habitats). These differences became smaller along a gradient of declining elevations. Where average groundwater levels were very high, around zero (very low elevation habitats of the recent floodplain), species showed the reverse behaviour. They tended to prefer lower groundwater levels, i.e. higher elevation habitats in the older floodplain. The relation between the optimum difference (urecent − uolder) and the gradient of urecent was strongly linear (F = 79·59 on 1 and 16 d.f., P < 0·0001).

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

The majority of plant species surveyed in this study exhibited significant responses to the water level fluctuation gradient. This was clearly demonstrated by the occurrence patterns related to the combination of average groundwater level/size of fluctuations and the patterns related to the floodplain type. Two major groups of species response were differentiated.

species shift along the elevation gradient

Most species tended to change their AGL optimum along the fluctuation gradient, especially those species preferring low AGL, i.e. species of high elevation habitats preferred much higher water levels in habitats with stable water tables. Considering the characteristics of fluctuating and stable water tables of the Elbe River floodplain (Fig. 1), there are significant differences at the beginning of the growing period, i.e. when grassland vegetation has a high requirement for soil moisture for growth (Gremmen et al. 1990; Gowing & Youngs 1997; Runhaar, Witte & Verburg 1997). The flow regime of the natural inundation area normally ensures sufficient soil moisture content during this time (the highest probability of high water is March–April). However, in the case of stable water tables, the groundwater level of high floodplain surfaces is too deep below the surface. Therefore, damping down water level fluctuations through the construction of dams and dykes causes drought stress at high elevations, which leads to species’ shifting towards much lower elevations. The shifts along the elevation gradient become smaller in mid- and low elevation habitats, because the higher groundwater level below the surface ensures a certain soil moisture content even in habitats of stable water levels. The situation is reversed in very low elevation areas such as floodplain depressions: species tended to shift from lower to higher elevations after water table stabilization. The explanation is well known from studies considering the effects of river regulation on river margin vegetation (Auble, Friedman & Scott 1994; Jansson et al. 2000; Nilsson & Berggren 2000): river margin habitats have a terrestrial phase within the yearly cycle in areas with highly fluctuating water tables, whereas they show a tendency to permanent inundation in areas with stable water tables. Terrestrial or semi-terrestrial species not adapted to water tables permanently above the surface have to colonize higher elevations after the onset of river regulation measures or else become extinct.

The implication of changes in the distribution of riparian vegetation after a reduction in water level fluctuations is evident: increased drought at higher elevation floodplains at one extreme, and an increased tendency to permanent inundation of lower elevation areas at the other extreme. This restricts the area that grassland species are able to colonize (Fig. 6) and therefore will cause a change in the spatial distribution of species. The two extremes will be colonized by drought and aeration stress-tolerant species, respectively. Similar patterns have been reported in studies comparing free-flowing and regulated rivers. Merritt & Cooper (2000) found that the vegetation of regulated streams with unnatural stable water flows in the Green River Basin, Colorado, USA, reflects a dichotomy in moisture conditions: marshes with anaerobic soils support wetland species, while terraces having xeric soil conditions support desert species. In contrast, the vegetation along unregulated rivers is characterized by a continuum of species distributed from the active channel to high floodplain areas. Similar results were reported by Auble, Friedman & Scott (1994), who studied the responses of riparian vegetation to alterations in hydrological regimes on the Gunnison River in Colorado, USA. The hypothetical stream flow, which assumed a stabilization of water levels without changing the mean annual flow, increased the dispersion of both the driest and the wettest vegetation types, i.e. it increased the area of vegetation types at the edges of the gradient that were either always or never inundated.

image

Figure 6. Shifts in species’ optima of AGL (circles) from highly fluctuating water tables (recent floodplain) towards stable conditions (older floodplain).

Download figure to PowerPoint

changes in species composition

Species composition varied between the floodplain types, with a number of species preferring either stable water tables on the older floodplain, or highly fluctuating water tables on the recent floodplain. This pattern has also been observed in other dyke-induced floodplain systems (Trémolières et al. 1998; Gergel, Dixon & Turner 2002). The preference for the older floodplain may be because these species require guaranteed levels of soil moisture while being simultaneously intolerant of floods. These conditions are not available in the recent floodplain, because only the highest elevations are rarely flooded and all the other habitats are characterized by long-lasting inundation and drought periods. For example, Holcus lanatus, a well-known flood-intolerant species (Müller-Stoll, Freitag & Krausch 1992; Hellberg 1995) of average moist habitats, is almost completely absent in the recent floodplain. Preference for the older floodplain may also be attributed to shallow flood depths. The recent floodplain often floods to a considerable depth during spring, which causes complete inundation of the plants. Therefore, early flowering plants, e.g. Cardamine pratensis, are at a considerable disadvantage in the recent floodplain even if they can tolerate long floods. Consequently, the main flowering time of grassland vegetation in the recent floodplain is mid-summer. However, in the older floodplain the reproductive parts of early flowering plants are not covered by water even in times of high water, because flooding depth hardly exceeds 50 cm above surface and is often much lower (Table 1). This may explain the frequent occurrence of Cardamine pratensis and other early flowering species (e.g. Rumex acetosa, Caltha palustris and Ranunculus acris) in this floodplain compartment.

Some species were significantly more common in the recent floodplain than in the older floodplain. The recent floodplain area is not only affected by a characteristic flow regime, but also by characteristic dispersal and disturbance processes. Hydrochory is important in structuring riparian flora because rivers carry large numbers of plant diaspores over considerable distances (Schneider & Sharitz 1988; Nilsson, Gardfjell & Grelsson 1991; Andersson, Nilsson & Johansson 2000) and river corridors are important pathways for plant dispersal (Malanson 1993; Johansson, Nilsson & Nilsson 1996). Furthermore, characteristic floodplain species often require the periodic availability of bare, wet surfaces and vegetation gaps for germination and recruitment, which strongly depends on flood-controlled disturbance events (Bornette & Amoros 1996; Bendix 1999). Floodplains fragmented by dykes and rivers regulated by dams have lost their function as dispersal pathways and do not provide disturbed, eroded and sedimented patches for plant establishment.

synthesis and applications

The general results of this study can be applied to other areas to predict the impacts of river regulation measures that restrict water level fluctuations. These measures can lead to considerable drought stress in higher elevation areas and permanent inundation at low elevation areas. This may be contradictory to the aim of river regulation, where the objective is to improve the agricultural use of the floodplain area. Complex drainage and irrigation systems may be necessary to compensate for drought and inundation in some areas. Restoration of floodplain meadows in these areas will fail, unless it includes the improvement of river–floodplain connectivity, for example by the relocation of dykes and by connecting former channels and other isolated water bodies to the main channel.

Polders are areas near the river surrounded by dykes that are artificially flooded by floodgates in times of extreme high water (often only a few times per century) to prevent the overflow and breach of dykes. This system is unsuitable for the maintenance of riparian vegetation because water level fluctuations are reduced in polders, as they are in the older floodplain areas, thereby limiting seed dispersal and natural disturbance processes. Furthermore, although the vegetation is radically affected by flood events, the frequency of occurrence is too low for vegetation to adapt. Thus, flood-control management, which has a high priority in Europe after significant flood events in recent years, should increasingly include dyke relocations and not just the construction of new polders.

Although the results from this study show clear and predictable effects of declining water level fluctuations, they are not directly transferable to other floodplain systems and to other regulation schemes. Large-scale factors include climate as well as the geological and geomorphological features of the catchment area (Naiman & Décamps 1990; Salo 1990). A meandering reach exhibits different responses to a given impact than a braided or constrained reach (Ward 1998), and the type of river regulation (e.g. by dams, dykes, diversion and canalization) also plays an important role. In this context, the time of flooding, for example a shift of seasonal high water from spring to summer and autumn, a common hydrological feature in storage reservoirs, has an impact independent of declining fluctuations (Nilsson, Jansson & Zinko 1997; Robertson, Bacon & Heagney 2001; Van Eck 2004).

Despite these limitations, the results from this study demonstrate the importance of water level fluctuations in determining the spatial and temporal distribution of floodplain species, as well as protecting the uniqueness of typical floodplain vegetation. The maintenance of dynamic water tables in addition to a high connectivity between river and floodplain is very important for the remaining unregulated and unfragmented river floodplain systems in central Europe. However, water level fluctuation is rarely used to characterize the ecological integrity of river floodplains. Additionally, intensity of floodplain fragmentation by dykes, defined, for example, as the ratio of recent to entire floodplain area, does not feature in assessment of the ecological status of riparian ecosystems (Palmer et al. 2005). The European Water Framework Directive (WFD; European Union 2000), which represents a fundamental reform of European Union (EU) water legislation in both environmental and administrative terms, making integrated river basin planning and management compulsory for member states, also pays little attention to the ecological status of floodplain systems (Lutosch, Scholz & Petry 2002). However, mitigation of the effects of river regulation measures and the successful restoration of floodplain ecosystems must focus on the re-establishment of river flow dynamics and the connectivity of the river with its floodplain. This requires an integrated approach in which the river floodplain system is considered as an ecological entity.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

I would like to thank all who helped me in the installation of the water level wells and the measurements of the water levels. I am grateful to the staff of the nature conservation administration in Brandenburg (Rühstädt) and Sachsen-Anhalt (Stendal) for their information and support. I thank Matthias Scholten for his advice on data analysis and two anonymous referees for helpful comments on the manuscript. Financial support was provided by the German Foundation for Environmental Issues (Deutsche Bundesstiftung Umwelt).

Supplementary material

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

Appendix. Results of logistic regression analyses for the subset ‘recent floodplain’ (r) and ‘older floodplain’ (o) for species in relation to AGL. AUC values and classification rates based on popt and pfair are displayed as criteria for the predictive power of the models.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information
  • Altman, D.G., Lausen, B., Sauerbrei, W. & Schumacher, M. (1994) Dangers of using ‘optimal’ cutpoints in the evaluation of prognostic factors. Journal of the National Cancer Institute, 86, 829835.
  • Andersson, E., Nilsson, C. & Johansson, M.E. (2000) Plant dispersal in boreal rivers and its relation to the diversity of riparian flora. Journal of Biogeography, 27, 10951106.
  • Auble, G.T., Friedman, J.M. & Scott, M.L. (1994) Relating riparian vegetation to present and future streamflows. Ecological Applications, 4, 544554.
  • Bendix, J. (1999) Stream power influence on southern Californian riparian vegetation. Journal of Vegetation Science, 10, 243252.
  • Bornette, G. & Amoros, C. (1996) Disturbance regimes and vegetation dynamics: role of floods in riverine wetlands. Journal of Vegetation Science, 7, 615622.
  • Ter Braak, C.J.F. & Looman, C.W.N. (1986) Weighted averaging, logistic regression and the Gaussian response model. Vegetatio, 65, 311.
  • Ter Braak, C.J.F. & Looman, C.W.N. (1995) Regression. Data Analysis in Community and Landscape Ecology (eds R.H.G.Jongman, C.J.F.Ter Braak & O.F.R.Van Tongeren), pp. 2966. University Press, Cambridge, UK.
  • Crawford, R.M.M. (1996) Whole plant adaptations to fluctuating water tables. Folia Geobotanica und Phytotaxonomica, 31, 724.
  • Crawley, M.J. (2003) Statistical Computing. An Introduction to Data Analysis Using S-Plus. Wiley, Chichester, UK.
  • Dahl, H.J. & Flade, M. (1994) Die Elbe und ihr Schutz: eine internationale Verpflichtung. Natur und Landschaft, 6, 239250.
  • Davis, M.M., Sprecher, S.W., Wakeley, J.S. & Best, G.R. (1996) Environmental gradients and identification of wetlands in north-central Florida. Wetlands, 16, 512523.
  • Deiller, A.F., Walter, J.N. & Trémolières, M. (2001) Effects of flood interruption on species richness, diversity and species composition of woody regeneration in the Upper Rhine alluvial hardwood forest. Regulated Rivers: Research and Management, 17, 393405.
  • Dister, E. (1991) Situation der Flussauen in der Bundesrepublik Deutschland. Erhaltung und Entwicklung Von Flußauen in Europa (ed. Bayerische Akademie für Naturschutz und Landschaftspflege), pp. 816. Laufener Seminarbeiträge No. 4. ANL, Laufen.
  • Dynesius, M. & Nilsson, C. (1994) Fragmentation and flow regulation of river systems in the northern third of the world. Science, 266, 753762.
  • European Union (2000) Directive 2000/60/EC of the European Parliament and the Council of 23 October 2000 establishing a framework for community action in the field of water policy. Official Journal of the European Communities, L327, 172.
  • Fielding, A.H. & Bell, J.F. (1997) A review of methods for the assessment of prediction errors in conservation presence–absence models. Environmental Conservation, 24, 3849.
  • Gergel, S.E., Dixon, M.D. & Turner, M.G. (2002) Consequences of human-altered floods: levees, floods and floodplain forests along the Wisconsin River. Ecological Applications, 12, 17551770.
  • Girel, J., Garguet-Duport, B. & Pautou, G. (1997) Landscape structure and historical processes along diked European valleys: a case study of the Arc/Isère Confluence (Savoie, France). Environmental Management, 21, 891907.
  • Gowing, D.J.G. & Spoor, G. (1998) The effect of water table depth on the distribution of plant species on lowland wet grassland. United Kingdom Floodplains (eds R.G.Bailey, B.V.José & B.R.Sherwood), pp. 185196. Westbury Academic and Scientific Publishing, Otley.
  • Gowing, D.J.G. & Youngs, E.G. (1997) The effect of the hydrology of a Thames flood meadow on its vegetation pattern. Floodplain Rivers: Hydrological Processes and Ecological Significance (ed. A.R.G.Large), pp. 6980. BHS Occasional Paper 8. British Hydrological Society, Newcastle-upon-Tyne.
  • Gremmen, N.J.M., Reijen, M.J.S., Wiertz, J. & Van Wirdum, G. (1990) A model to predict and assess the effects of groundwater withdrawal on the vegetation in the Pleistocene areas of the Netherlands. Journal of Environmental Management, 31, 143155.
  • Hanley, J.A. & McNeil, B.J. (1982) The meaning and use of the area under a ROC curve. Radiology, 143, 2936.
  • Hellberg, F. (1995) Entwicklung der Grünlandvegetation bei Wiedervernässung und periodischer Überflutung. Dissertationes Botanicae, 243, 1271.
  • Hill, N.M., Keddy, P.A. & Wisheu, I.C. (1998) A hydrological model for predicting the effects of dams on the shoreline vegetation of lakes and reservoirs. Environmental Management, 22, 732736.
  • Hosmer, D.W. & Lemeshow, S. (2000) Applied Logistic Regression. Wiley, New York, NY.
  • Huberty, C.J. (1994) Applied Discriminant Analysis. Wiley, New York, NY.
  • Jansson, R., Nilsson, C. & Renöfält, B. (2000) Fragmentation of riparian floras in rivers with multiple dams. Ecology, 81, 899903.
  • Jansson, R., Nilsson, C., Dynesius, M. & Andersson, E. (2000) Effects of river regulation on river margin vegetation: a comparison of eight boreal rivers. Ecological Applications, 10, 203224.
  • Johansson, M.E. & Nilsson, C. (2002) Responses of riparian plants to flooding in free-flowing and regulated boreal rivers: an experimental study. Journal of Applied Ecology, 39, 971986.
  • Johansson, M.E., Nilsson, C. & Nilsson, E. (1996) Do rivers function as corridors for plant dispersal? Journal of Vegetation Science, 7, 593598.
  • Johnston Associates (1989) A Status Report on the Nation's Floodplain Management Activity: An Interim Report Prepared for Interagency Task Force on Floodplain Management. US Government Printing Office, Washington, DC.
  • Leyer, I. (2002) Auengrünland der Mittelelbe-Niederung: Vegetationskundliche und ökologische Untersuchungen in der rezenten Aue, der Altaue und am Auenrand der Elbe. Dissertationes Botanicae, 363, 1193.
  • Leyer, I. (2004) Effects of dykes on plant species composition in a large lowland river floodplain. River Research and Applications, 20, 813827.
  • Londo, G. (1976) The decimal scale for relevees of permanent quadrats. Vegetatio, 33, 6164.
  • Lutosch, I., Petry, D. & Scholz, M. (2002) Auen und Auenschutz in der Wasserrahmenrichtlinie. Relevanz der EU Wasserrahmenrichtlinie für Den Naturschutz in Auen (eds D.Petry, M.Scholz & I.Lutosch), pp. 1042. UFZ-Bericht 22, Leipzig.
  • Malanson, G.P. (1993) Riparian Landscapes. Cambridge University Press, Cambridge, UK.
  • Marston, R.A., Girel, J., Pautou, G., Piegay, H., Bravard, J. & Arneson, C. (1995) Channel metamorphosis, floodplain disturbance, and vegetation development: Ain River, France. Geomorphology, 13, 121131.
  • Merritt, D.M. & Cooper, D.J. (2000) Riparian vegetation and channel change in response to river regulation: a comparative study of regulated and unregulated streams in the Green River Basin, USA. Regulated Rivers: Research and Management, 16, 543564.
  • Müller-Stoll, W.R., Freitag, H. & Krausch, K.H. (1992) Die Grünlandgesellschaften des Spreewaldes. Gleditschia, 20, 273302.
  • Naiman, R.J. & Décamps, H. (1990) Aquatic–terrestrial ecotones: summary and recommendations. The Ecology and Management of Aquatic–Terrestrial Ecotones (eds R.J.Naiman & H.Décamps), pp. 295301. UNESCO, Paris.
  • Nilsson, C. & Berggren, K. (2000) Alterations of riparian ecosystems caused by river regulation. Bioscience, 50, 783792.
  • Nilsson, C. & Jansson, R. (1995) Floristic differences between riparian corridors of regulated and free-flowing boreal rivers. Regulated Rivers: Research and Management, 11, 5566.
  • Nilsson, C., Gardfjell, M. & Grelsson, G. (1991) Importance of hydrochory in structuring plant communities along rivers. Canadian Journal of Botany, 69, 26312633.
  • Nilsson, C., Jansson, R. & Zinko, U. (1997) Long-term responses of river-margin vegetation to water level regulation. Science, 276, 798800.
  • Palmer, M.A., Bernhardt, E., Allan, D., Lake, P.S.S. & NRRSS (2005) Standards for ecologically successful river restoration. Journal of Applied Ecology, 42, DOI: 10.1111/j.1365-2664.2005.01004.x.
  • Petts, G.E. (1984) Impounded Rivers. Wiley, Chichester, UK.
  • Petts, G.E. (1989) Historical analysis of fluvial hydrosystems. Historical Change of Large Alluvial Rivers: Western Europe (eds G.E.Petts, H.Möller & A.L.Roux), pp. 118. Wiley, Chichester, UK.
  • R-Development Core Team (2004) R: A language and Environment for Statistical Computing. http://www.R-project.org.
  • Robertson, A.I., Bacon, P. & Heagney, G. (2001) The response of floodplain primary production to flood frequency and timing. Journal of Applied Ecology, 38, 126136.
  • Runhaar, H., Witte, F. & Verburg, P. (1997) Ground-water level, moisture supply and vegetation in the Netherlands. Wetlands, 17, 528538.
  • Salo, J. (1990) External processes influencing origin and maintenance of inland water-land ecotones. The Ecology and Management of Aquatic-Terrestrial Ecotones (eds R.J.Naiman & H.Décamps), pp. 3764.
  • Schneider, E. (1991) Die Auen im Einzugsgebiet der unteren Donau. Erhaltung und Entwicklung Von Flußauen in Europa (ed. Bayerische Akademie für Naturschutz und Landschaftspflege), pp. 4057. Laufener Seminarbeiträge No. 4. ANL, Laufen.
  • Schneider, R.L. & Sharitz, R.R. (1988) Hydrochory and regeneration in a bald cypress–water tupelo swamp forest. Ecology, 69, 10551063.
  • Schröder, B. (2000) Zwischen Naturschutz und Theoretischer Ökologie: Modelle zur Habitateignung und räumlichen Populationsdynamik für Heuschrecken im Niedermoor. Selbstverlag Institut für Geographie und Geoökologie der Technischen Universität Braunschweig, Braunschweig, Germany.
  • Schröder, B. (2004) ROC_AUC: Evaluating the Predictive Performance of Habitat Models. http://brandenburg.geoecology.uni-potsdam.de/users/schroeder/download.html.
  • Schröder, B. & Richter, O. (1999) Are habitat models transferable in space and time? Zeitschrift für Ökologie und Naturschutz, 4, 195206.
  • Sokal, R.R. & Rohlf, F.J. (1995) Biometry. Freeman, New York, NY.
  • Thomas, J.A. & Bovee, K.D. (1993) Application and testing of a procedure to evaluate transferability of habitat suitability criteria. Regulated Rivers: Research and Management, 8, 285294.
  • Tockner, K. & Stanford, J.A. (2002) Riverine flood plains: present state and future trends. Environmental Conservation, 29, 308330.
  • Trémolières, M., Sánchez-Pérez, J.M., Schnitzler, A. & Schmitt, D. (1998) Impact of river management history on the community structure, species composition and nutrient status in the Rhine alluvial hardwood forest. Plant Ecology, 135, 5978.
  • Van Eck, W. (2004) Flooding tolerance and the distribution of plant species along flooding gradients. PhD Thesis. University of Nijmegen, Nijmegen, the Netherlands.
  • Ward, J.V. (1998) Riverine landscapes: biodiversity patterns, disturbance regimes, and aquatic conservation. Biological Conservation, 83, 269278.
  • Wisskirchen, R. & Haeupler, H. (1998) Standardliste der Farn- und Blütenpflanzen Deutschlands. Ulmer, Stuttgart, Germany.
  • Zweig, M.H. & Campbell, G. (1993) Receiver-operating characteristic (ROC) plots: a fundamental evaluation tool in clinical medicine. Clinical Chemistry, 39, 561577.
  • Zweig, M.H. & Robertson, E.A. (1982) Why we need better test evaluations. Clinical Chemistry, 28, 12721276.

Received 15 March 2004; final copy received 15 December 2004

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

Appendix S1. Results of logistic regression analyses for the subset ?recent floodplain? (r) and older floodplain (o) for species in relation to AGL. AUC values and classification rates based on popt and pfair are displayed as criteria for the predictive power of the models.

FilenameFormatSizeDescription
JPE_1009_sm_Appendix S1.xls31KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.