SEARCH

SEARCH BY CITATION

Keywords:

  • acidification;
  • atmospheric deposition;
  • ecological amplitude;
  • eutrophication;
  • plant species conservation;
  • Red Data Book species

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
  • 1
    During the last century, many plant species typical of heathland and nutrient-poor acidic grasslands have become rare whereas others have remained common. Habitat restoration often fails to enhance the rare species, which may in part be caused by the failure to restore the biogeochemical conditions suited to these species. Many soil variables have been shown to affect plant fitness but it is unknown what their relative importance is and whether any biogeochemical variable acts as a key factor constraining the persistence of rare heathland species.
  • 2
    We compiled a data set consisting of 300 vegetation samples and the associated soil chemical properties from a range of studies carried out across the Netherlands. We asked whether growth sites of rare and common species typical of heathland and acidic grasslands differed in their biogeochemical properties, and whether growth sites of rare species displayed less variation in soil biogeochemical variables (e.g. had narrower ecological amplitude).
  • 3
    Regardless of rarity, the species’ growth sites were most accurately described by a curvilinear relationship between pH and Al/Ca ratios. Other soil characteristics did not vary systematically with changing acidity of the soil or the patterns were less pronounced. Acidification will therefore most rapidly and predictably result in an increase in Al/Ca ratio whereas this is not necessarily the case for the other soil variables affecting plant fitness.
  • 4
    The soil ammonium (NH4) concentration and ammonium/nitrate (NH4/NO3) ratio were 3·5 and 3 times higher, respectively, in growth sites of common species compared with those of rare species. No other measured variable differed significantly between rare and common plant species.
  • 5
    On average rare species had a significantly narrower ecological amplitude than common species for soil biogeochemical parameters.
  • 6
    Synthesis and applications. A greater sensitivity to high NH4 concentrations in combination with a narrower ecological tolerance zone for a range of soil biogeochemical factors may explain the demise of rare species adapted to nutrient-poor acidic habitats in recent decades. Conservation management should aim to restore low NH4 concentrations and NH4/NO3 ratios. Experimental studies indicate that the most effective way to do this is through removal of the topsoil in combination with liming.

Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Heathlands, matgrass swards and fen meadows are human-made habitats. Located on nutrient-poor, often acidic, soils, they were used by farmers for grazing, hay-cutting, turf-cutting and collection of firewood. Often more nutrients were removed from the system than were returned to it as animal excrements, atmospheric deposition or fertilizer (Spek 2005). This created a suitable and often exclusive habitat for a range of characteristic plant species adapted to these stressful conditions, and created habitats of conservation interest even though they are not particularly species-rich. Heathlands and associated acidic grasslands were once common throughout the Atlantic regions of western Europe (Webb 1998). During the last century, they have declined significantly both in area and in quality because of conversion to more productive types of agricultural land use, changes in or abandonment of management practices, regional hydrological changes and atmospheric deposition (Bobbink, Hornung & Roelofs 1998; Webb 1998; Jansen, Eysink & Maas 2001; Vandvik et al. 2005). As a result, the European Union Habitats Directive (Directive 92/43/EEC) now lists heathland amongst the habitat types whose conservation requires the designation of special areas of conservation.

Conservation initiatives consist of the legal protection of existing areas of heathland and acidic grasslands, the restoration of these habitats on ex-agricultural land (Walker et al. 2004) and the implementation of agri-environment schemes on privately owned land (Marrs et al. 2004). Despite the return to traditional management practices and extensive restoration measures (e.g. rewetting and removal of the nutrient-rich topsoil), typical species of heathlands and related grasslands often fail to establish or increase in population size (Bekker & Lammerts 2000; Walker et al. 2004). In part, this may be related to the poor dispersability and short seed longevity of the more endangered heathland species (Strykstra, Pegtel & Bergsma 1998; Matus et al. 2003). Species will fail to colonize sites when they have disappeared from the seed bank and when the nearest source population is beyond dispersal range. In part, it may also be related to the fact that conservation management fails to restore the biogeochemical conditions suited to plant species adapted to acidic nutrient-poor habitats. This study focused on the second aspect and explored in what way the conservation of endangered species is constrained by the abiotic characteristics of heathlands.

The biogeochemical properties of heathland soils are closely related to soil pH, which may vary from 3 to 6. During the last century, pH in most acidic nutrient-poor habitats declined significantly as a result of atmospheric deposition of sulphur (S) and nitrogen (N) compounds (Van Breemen & Wright 2004). This acidification may be buffered by the exchange of the base cations for protons (H+) at the soil absorption complex (Fig. 1a). Subsequent leaching of base cations may eventually lead to a decline in soil pH and a shift to aluminium (Al3+) buffering. This results in higher concentrations of the potentially phytotoxic substances aluminium and ammonium (inline image; Britto & Kronzucker 2002; Kinraide 2003) in the soil volume (Fig. 1b). So far the demise of heathland species and the failure of many heathland restoration projects has been attributed to the sensitivity of these species to eutrophication (Roelofs 1986), acidification (Houdijk et al. 1993), high soil inline image concentrations (Van den Berg et al. 2005), high soil ammonium/nitrate inline image ratios (Van den Berg 2006), high soil Al3+ concentrations and high soil aluminium/calcium (Al3+/Ca2+) ratios (Fennema 1992; Houdijk et al. 1993). However, most studies have examined the effects on only a few species or focused on one or two soil chemical properties. This makes it impossible to generalize and it is currently unknown what the relative importance of the different factors is whether this varies between species and whether threshold levels can be identified beyond which it is impossible for species to persist.

image

Figure 1. The predominant biogeochemical processes affecting plant growth at acidic (pH 4·5–5·5) and acid (pH 3·5–4·5) growth sites. (a) In acidic soils acidification may occur because of deposition of SOx compounds, nitrification and uptake of cations. Acidification is counteracted by denitrification, which consumes protons (H +) and is buffered by the exchange of base cations (Ca2+, Mg2+, K+; indicated by Ca2+ in the graph) at the soil absorption complex. Base cations may subsequently leach from the rooting zone, resulting in base cation depletion. (b) In acid soils base cations are depleted and at the soil absorption complex H+ is exchanged for Al3+, resulting in increased Al3+ concentrations in the soil volume. At low pH nitrification is inhibited (Roelofs et al. 1985), resulting in reduced denitrification and inline image accumulation.

Download figure to PowerPoint

We used the original data from a range of studies across the Netherlands that linked the vegetation composition of heathlands and acidic grasslands with a range of soil chemical properties. A species may be absent because of chance or failure to disperse to an otherwise suitable site; the absence of a species cannot be used as an indication of the biogeochemical unsuitability of a particular site. Therefore we only used data from sites occupied by species of interest. For reference we compared the response of species that have declined significantly over the past decades and are listed in the Dutch Red Data Book (Van der Meijden et al. 2000; henceforth ‘rare’ species) with species that have not significantly declined, assuming that the difference in rarity is somehow linked to plant traits such as tolerance to biogeochemical conditions (Kunin & Gaston 1993). We addressed the following research questions. Do the growth sites of rare and common species differ in their biogeochemical properties? Do rare and common species differ in the range of biogeochemical parameters in their growth sites (ecological amplitude)? Ultimately, we aimed to formulate a set of simple recommendations or guidelines that may be used to enhance the success of conservation initiatives for heathland ecosystems.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

data collection

We compiled a data set containing 300 vegetation samples and associated soil biogeochemical properties. Criteria for inclusion of samples in the data set were the presence of at least one species indicative of dry and wet heathland (vegetation types Calluno–Genistion pilosae and Ericion tetralicis, respectively), matgrass swards (Nardo–Galion saxatilis) or fen meadows (Junco–Molinion) and data on at least soil pH and concentrations of Ca2+, Al3+, inline image and preferably also S2–, phosphate inline image, base cations (BC), base saturation (BS) and cation exchange capacity (CEC). Data came from Houdijk et al. (1993), De Graaf et al. (1994) and a number of unpublished studies all carried out in the period 1986–2003 at the Department of Aquatic Ecology and Environmental Biology of Nijmegen University (Nijmegen, the Netherlands). The data were collected in 87 different areas, mostly nature reserves in the eastern, central and southern parts of the Netherlands. It proved impossible to obtain management information from all these sites so, for the purpose of this study, we assumed that differences in management were reflected in differences in soil variables. The large majority of the samples were taken in areas with relatively species-rich vegetation. The potential bias that may arise from this is highlighted in the Discussion. The number of samples per area varied between 1 and 44 but the majority of the areas was represented with just one or two samples. The size of the plots used to sample the vegetation varied between studies and areas but was on average 29 m2 (SE 6·5 m2).

soil sampling and chemical analyses

In each vegetation sample the upper 10 cm of the soil was sampled using a 3-cm diameter auger. Eight to 12 subsamples were pooled, mixed and stored in polyethylene bags at 4 °C until analysis.

Seventy grams of fresh soil were mixed with 200 mL bidistilled water (for determination of water-extractable elements) or 200 mL 0·2 m NaCl solution (exchangeable elements). The mixtures were shaken for 1 h (120 movements min−1), after which the pH of the solution was measured (radiometer type PHM 82 pH-meter; Radiometer, Denmark). Subsequently, the solution was centrifuged (12 000 r.p.m., 20 min) and the supernatant stored in polyethylene bottles at –28 °C. Al, Ca, magnesium (Mg2+), manganese (Mn2+), iron (Fe2+), zinc (Zn2+) and S concentrations were determined using an ICP (Inductively coupled plasma) optical emission spectrometer (Jarrell Ash Plasma-200; Jarrell Ash, Grand Junction, Co, USA). K was analysed with flame photometry (Technicon Flame Photometer IV; Technicon, Tarrytown, NY, USA). inline image concentrations were determined colorimetrically with a continuous-flow autoanalyser (Technicon AAII system; Technicon, Tarrytown, NY, USA). Concentrations of all elements were expressed in µmol kg dry soil−1.

The BC concentration was calculated as the exchangeable concentration (in terms of charge equivalents, concentration multiplied by the charge of the molecules) of the base cations Ca, Mg and K. CEC was calculated as (charge concentration of all exchangeable cations) – (charge concentration of all water-extractable cations)/1000. BS was calculated as the percentage exchangeable BC of the total exchangeable cations also expressed in charge concentrations.

analysis

Initially, we selected 37 species typical of heathland and acidic grasslands. For each of these species we subsequently selected all samples containing the species, regardless of its abundance. To avoid pseudoreplication, individual areas never contributed more than two randomly selected samples. Samples usually contained more than one species of interest and generally contributed to the calculation of means of all species present in them. We had insufficient samples for 10 species: Antennaria dioica, Carex oederi, Carex pulicaris, Epipactis palustris, Genista tinctoria, Nardus stricta, Parnassia palustris, Thymus serpyllum, Veronica scutellata and Viola canina. All but Carex oederi and Veronica scutellata were Red Data Book species (nomenclature following Van der Meijden 1990). Of the remaining 27 species, 16 were rare and 11 were common (Table 1).

Table 1.  Mean values of some key biogeochemical variables measured at growth sites of species characteristic of dry and wet heathland, matgrass swards and fen meadows. NH4 was extracted in 0·2 m NaCl solution, all other elements were extracted in H2O. pH is pH (H2O). All units are in µmol kg dry soil−1 except BC (µeq kg dry soil−1), BS (% of all cations) and CEC (µeq kg dry soil−1); NA, not available. Test statistics are given for differences between common and rare species after effects of pH had been accounted for
 pHAlCaAl/CaNH4NO3NH4/NO3SPO4BCBSCEC
  • *

    25 d.f.,

  • †22 d.f.,

  • ‡20 d.f.

Rare species
Arnica montana4·49188·160·05·97164·613·612·83145·898·333 82525·312·2
Cirsium dissectum5·39243·7988·50·46339·166·513·81944·71NA70 21086·718·3
Dactylorhiza maculata5·02363·5482·51·41164·233·511·56375·7512·5328 21579·623·6
Drosera intermedia4·56190·8119·34·5467·312·07·98194·255·003 93433·78·7
Drosera rotundifolia4·66176·0117·93·69166·310·441·41177·646·434 22434·08·2
Genista anglica4·54309·4100·96·06195·022·66·51191·3811·507 83042·815·9
Genista pilosa4·37138·248·310·07240·410·440·57190·783·562 49630·18·5
Gentiana pneumonanthe4·74338·7355·32·66139·837·17·88348·7315·3117 98160·123·2
Lycopodium clavatum4·05214·638·87·8976·827·93·80191·627·541 84813·212·7
Lycopodium inundatum4·84142·4134·51·7897·78·315·70198·233·903 95842·85·5
Narthecium ossifragum4·79221·2253·73·92220·627·511·24726·5015·6718 49827·114·9
Pedicularis sylvatica5·13389·2426·72·69174·016·412·17268·447·3626 03671·520·9
Polygala serpyllifolia4·69327·2362·22·44165·725·36·58338·8012·6414 98452·018·3
Rhynchospora alba4·64163·2185·26·3351·58·611·43165·947·334 75928·88·3
Rhynchospora fusca4·38234·4135·25·74145·219·812·65336·068·297 67826·710·9
Succisa pratensis5·14295·2503·91·29523·544·69·69299·857·1732 56270·725·7
Common species
Calluna vulgaris4·48180·3124·66·68149·215·315·87194·497·305 81736·511·1
Carex echinata5·16105·8376·00·59100·311·633·96NANA14 44683·421·1
Carex nigra5·00297·2389·02·34531·327·816·91337·958·8927 41074·424·2
Carex panicea5·18270·0669·41·20831·652·865·20641·2114·5842 73363·828·2
Danthonia decumbens4·81208·8215·33·37565·932·327·53251·429·4512 92356·216·2
Erica tetralix4·55246·7237·04·06339·527·016·16368·0211·7811 02342·313·9
Eriophorum angustifolium5·07192·3642·21·691183·5105·9103·03NANA26 907NANA
Juncus acutiflorus5·22249·5531·50·96514·927·341·56453·7912·5529 51967·818·8
Molinia caerulea4·60264·1381·05·29322·843·513·52496·2010·8321 27732·013·5
Potentilla erecta4·83277·1405·83·22619·441·254·37308·7911·9419 66148·817·5
Viola palustris5·61191·51264·10·161989·3144·779·55969·98NA92 834NANA
Mean rare species4·71246·0269·64·18183·224·014·11318·418·8415 56345·314·7
Mean common species4·95225·7476·02·69649·848·142·51446·8710·9227 68656·118·3
t241·82*1·10–0·64–0·03–2·62–1·28–2·82–0·54–1·02–0·06–0·45–0·81
P0·0810·2840·5270·9740·0150·2100·0100·6000·321 0·9540·6560·425

In a first step, we determined mean values of all soil variables for each species. Because the different studies that produced the basic data for this analysis differed somewhat in the range of soil variables measured, we did not have data for all soil variables in all samples. The basic number of areas and samples used to calculate the means of each species is given in Table S1 in the supplementary material but these may be lower for PO4, S, BS and CEC because they were not measured in all samples. Means were not included in the analyses or presented in tables if they were based on less than five samples. Overall sample size for soil variables was 16 ± 0·4 (mean ± SE) and 35 ± 1·7 samples for rare and common species, respectively.

To examine whether rare and common species differed in the range of biogeochemical conditions they can tolerate, we calculated a coefficient of variation (CV) for each species and soil variable combination. CV is the standard deviation of observations divided by the mean of observations so that comparisons between different variables are possible.

Although various species are considered typical of certain vegetation types (e.g. dry heaths and matgrass swards), most species do occur in more than one vegetation type and some in all (Schaminée, Stortelder & Weeda 1996). As the focus of this study was on the differences in the level and range of soil variables that can be tolerated by species, we decided to analyse the species data regardless of the vegetation type where species have their optimum distribution.

Our main interest was to examine whether there were differences in soil characteristics of sites with rare and common species. For all statistical tests we considered the different species as replicates of the ‘treatment’ rarity and performed our analyses on the species’ means or CV for each soil variable. The pH of the growth sites of endangered species and common species differed somewhat (Table 1) and pH affected the concentration of most of our variables of interest. We therefore analysed the data using a covariance analysis with rarity as the main factor and pH as the covariance factor. We tested whether differences in the means of the CV differed between common and endangered species using unpaired t-tests.

On average there was a 13% overlap between samples used to calculate means of one species with samples used to calculate means of other species. A partial overlap was inevitable because the species of interest often grow together. Using only sites where single species occur would probably overestimate differences between species. Statistically, this overlap may pose a problem because tests assume independence of observations and overlap may result in a reduced variance in the data set, resulting, in turn, in progressive statistical tests. A simulation study was done to quantify how the study design affected the outcome of the analyses. For pH and inline image, mean and variance were estimated using the described site selection procedure (e.g. maximum of two randomly selected plots per area). Each sample was subsequently allocated a random value drawn from the normal distribution based on this mean and variance. A covariance analysis was then performed 5000 times with a statistical error (α) of 5% to test whether the mean of rare species was significantly different from that of common species. A proper test should result in 5% of the resulting P-values being lower than 0·05, indicating rejection of the null hypothesis. The simulation resulted in a rejection of slightly less then 5% of the cases, indicating that the test was slightly conservative. We therefore believe our approach was suitable for the objective of this study. Where necessary, the data were log-transformed to achieve normality and constant variance of the response variables. All analyses were carried out using the statistical package genstat (Payne et al. 2002).

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Growth sites of rare and common species differed in the concentration of soil NH4 and the NH4/NO3 ratio (Table 1). The significantly lower mean NH4 concentrations and NH4/NO3 ratio were not influenced by pH. NH4/NO3 ratios tended to increase with increasing pH but only for common species (Fig. 2). Regardless of pH, the mean NH4/NO3 ratio at the growth sites of rare species did not exceed 16 except for Drosera rotundifolia and Genista pilosa, both of which had atypically high standard errors caused by one or two growth sites with NH4/NO3 ratios that were 4–6 times higher than the next highest ratio. NH4/NO3 ratios in growth sites of common species were all higher than 13 (Table 1).

image

Figure 2. Characterization of growth sites of common (closed diamonds) and rare (open squares) species typical of heathland, matgrass swards and fen meadows by means of pH and NH4/NO3 ratio. Symbols indicate mean ± SE.

Download figure to PowerPoint

Mean values of all other measured soil variables varied considerably between growth sites of different species and did not differ systematically between rare and common species. Contrary to expectation, the Al concentration and Al/Ca ratio were slightly higher and Ca concentration lower at growth sites of rare species compared with common species. This was primarily because of the lower pH of growth sites of rare species (Table 1) and the strong curvilinear relationships between mean pH and mean Al/Ca ratio and Ca concentration at the growth sites of the 27 heathland species (Fig. 3). Although rare and common species occupied slightly different positions, species from both groups could be found at the high and low end of the pH–Al/Ca ratio gradient, precluding systematic differences in Al/Ca ratio. These patterns in mean soil properties of individual species simply reflected soil conditions of the examined sites. In our data set both Ca concentrations and Al/Ca ratios were related to pH independent of the type of species growing at those sites [log(Ca concentration) = 1·30 × pH – 1·53, F1,117 = 46,96, P < 0·001; log(Al/Ca ratio) = 8·08–1·60 × pH, F1,117 = 124,7, P < 0·001]. Combinations of pH and other soil properties at growth sites of heathland species showed less pronounced patterns.

image

Figure 3. Characterization of growth sites of common (closed diamonds) and rare (open squares) species typical of heathland, matgrass swards and fen meadows by means of pH and (a) Al/Ca ratio or (b) Ca concentration (µmol kg dry soil−1). The mean values of soil parameters of different species display predictable patterns described by log(Al/Ca-ratio) = 12·99–2·50 × pH (F1,25 = 140·5, P < 0·001, adjusted R2 = 84%) and log(Ca concentration) = 2·25 × pH – 5·29 (F1,25 = 96·7, P < 0·001, adjusted R2 = 79%). Symbols indicate mean ± SE.

Download figure to PowerPoint

For all soil variables the CV was lower for rare species than for common species (Table 2). Compared with the growth sites of common species, those of rare species displayed significantly less variation in soil concentrations of H+, Ca and BC and in CEC. As a result the mean CV for all soil variables was significantly lower for rare species. The basic pattern of variation across soil variables was the same for common and rare species; both groups displayed low variation for S and Al concentrations and CEC and high variation for NH4, Ca and BC concentrations.

Table 2.  The CV of some key biogeochemical variables measured at growth sites of species characteristic of dry and wet heathland, matgrass swards and fen meadows. H+ concentrations are given instead of pH [i.e. –log(H+)] so that H+ CV can be compared with that of other elements. mCV, mean coefficient of variation of all variables. BS is expressed as a percentage thus reducing variation and CV
 HAlCaAl/CaNH4NO3NH4/NO3SPO4BCBSCECmCV
Mean rare species0·830·740·990·881·070·830·920·670·780·950·630·640·83
Mean common species1·090·821·491·171·240·971·300·770·831·340·630·811·04
t212·880·912·821·910·821·211·700·630·942·710·052·823·40
P0·0090·3710·0100·0700·4230·2400·1050·5360·3570·0250·9650·0100·003

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

None of the species included in this study are endemic or rare by nature. Most of the species that are now classified as rare used to be quite common in the Netherlands and many are still common in neighbouring countries. Some species have become rare following large-scale habitat destruction and changes in environmental conditions in their remaining habitats, whereas others have remained fairly common. The current comparison of biogeochemical conditions experienced by common and rare species in remaining heathlands, matgrass swards and fen meadows suggests that a higher sensitivity to NH4 and a relatively low biogeochemical niche breadth may have played an important role in the decline of rare species in the Netherlands.

do the growth sites of endangered and common species differ in their biogeochemical properties?

Growth sites of rare and common plant species typical of heathland and closely related acidic nutrient-poor grasslands differed significantly in the concentration of soil NH4 and the soil NH4/NO3 ratio but not in concentrations of any other biogeochemical factor known to influence plant fitness. The absence of rare species from sites with high concentrations of NH4 may be the result of them being outcompeted by more competitive species. Molinia caerulea, in particular, has been found to benefit from increased levels of N and may outcompete most other heathland species at high N availability (Aerts & Berendse 1988; Bobbink, Hornung & Roelofs 1998). Additionally, in experimental studies a range of the investigated rare species has been found to be more sensitive to high NH4 concentrations or high NH4/NO3 ratios (Dueck & Elderson 1992; De Graaf et al. 1998a; Van den Berg et al. 2005; Van den Berg 2006). The fact that common species with low competitive ability, such as Potentilla erecta and Danthonia decumbens, persisted in soils with considerably higher NH4 concentrations and NH4/NO3 ratios than those measured for Molinia caerulea suggests that differences in sensitivity to NH4 toxicity do contribute to the absence of rare species from sites with high NH4 concentrations.

N proved to be a better predictor of rare plant species occurrence than P, as soil PO4 concentrations did not differ significantly between stands of common and rare species. Wassen et al. (2005) suggested that P rather than N limitation was particularly important for the persistence of endangered species in herbaceous ecosystems because most of them occurred at plant N/P ratios > 16, indicative of P-limited conditions (Güsewell & Koerselman 2002). Assuming that the soil-available N/P ratios in our study are indicative of the type of nutrient limitation, common species grew under P-limited conditions but rare species grew under N-limited conditions (< 13·5; Güsewell & Koerselman 2002; mean ± SE, 21·0 ± 2·6 vs. 11·7 ± 2·4 for common and rare species, respectively, F1,19 = 6·77, P= 0·018). Güsewell & Koerselman (2002) pointed out that N/P ratios in vegetation and in soil are often poorly correlated and soil N/P ratios may not be as indicative of the type of nutrient limitation as plant N/P ratios. In any case, enhanced N availability caused by atmospheric deposition of N compounds will result in less N-limited conditions and more P-limited conditions. Our findings show that, for acidic nutrient-poor ecosystems, a shift in this direction is associated with a decline in rare species. The contrasting results of Wassen et al. (2005) may have been caused by their study being restricted to wet herbaceous habitats and their findings may not extend to dry or even moist herbaceous ecosystems. The reserves investigated in this study all had very low soil P concentrations (Janssens et al. 1998). In heathland restoration projects on former agricultural land, P concentrations are generally much higher and may pose a serious impediment to the establishment of rare heathland species.

The growth sites of species of acidic nutrient-poor habitats segregated most predictably along a gradient of pH and Al/Ca ratios (Fig. 3a). Other soil characteristics did not vary systematically with changing acidity of the soil or the patterns were less pronounced. Acidification will therefore most rapidly and predictably result in an increase in Al/Ca ratios whereas this is not necessarily the case for other examined variables. No systematic differences were observed between the Al concentrations or Al/Ca ratios at the growth sites of common and rare species, but individual species did occur on sites covering a restricted range of pH (and associated Al/Ca ratios). Acidification may therefore result in the habitat becoming unsuitable for both common and rare species. However, because rare species have a significantly smaller tolerance zone for acidity (H+; Table 2), they are likely to be more adversely affected. Because of the strong relationship between the pH and Al/Ca ratio it is difficult to establish which factor is more important in constraining the persistence of species. Furthermore, this suggests that the relationship between Al/Ca ratios and the occurrence of rare heathland species observed by Fennema (1992) and Houdijk et al. (1993) could equally well be explained by differences in pH rather than differences in tolerance to high Al/Ca ratios.

A potential bias may have been caused by eutrophication and acidification having more severe adverse effects on recruitment than on growth and survival of established plants (Vergeer et al. 2003). Because many of the examined species are long-lived there may be a time-lag between deterioration of habitat quality and local extinction of species. In this study we could not distinguish between viable populations and populations consisting of a few old plants hanging on narrowly. This study may therefore overestimate the range in biogeochemical variables suitable for rare species. A second bias may have arisen from that fact that most samples were taken in remnants of species-rich heathland and few samples were taken in species-poor heathland dominated by Calluna vulgaris or Molinia caerulea. In these sites biogeochemical conditions may have been considerably more inhospitable to rare species. Both potential biases may have led to an underestimation of the difference between common and rare species and results should be interpreted bearing this in mind.

do endangered and common species differ in the biogeochemical range they tolerate?

Rare species had, on average, a significantly narrower ecological amplitude than common species for soil biogeochemical parameters. CV tend to increase with decreasing sample sizes and the sample size of rare species was considerably smaller than those of common species, therefore the observed differences are probably conservative estimates. The observed differences could have been caused by the different phylogenetic background of rare and common species and because most rare species were subordinate species whereas most common species were (sub)dominants. However, a comparison of the CV of the common Carex species with the rare Rhynchospora species, all members of the Cyperaceae family and all (sub)dominant species, confirmed the general pattern (average mCV of 0·81 and 0·57 for Carex and Rhynchospora species, respectively). Regardless of why the examined species differ in their niche breadth, it is a trait that probably contributed to their recent decline in north-western Europe and our findings support the hypothesis that a reduced niche breadth, or reduced physiological tolerance zone, could restrict the distribution and abundance of species (Brown 1984).

Interestingly, individual factors for which CV differed significantly between rare and common species were those related to the buffering capacity of the habitat (Ca, BC and CEC) rather than potentially toxic or eutrophicating factors. We do not have a satisfying explanation for this result.

synthesis and applications

North-western European heathland habitats are continuously being acidified and eutrophicated (Pearson & Stewart 1993; van Breemen & Wright 2004). The significantly narrower ecological tolerance zone for soil biogeochemical factors of rare species probably explains why these species have become listed in the Dutch Red Data Book rather than the co-occurring species that are still relatively common. Following the gradual change in biogeochemical conditions of their growth sites, the tolerance thresholds for these species were probably exceeded more rapidly than those of common species, which, in combination with a lower reproductive capacity and dispersability (Bekker & Kwak 2005), resulted in a more pronounced decline.

Although our study does not rule out adverse effects of other soil variables, it does suggest that NH4 has an overriding adverse effect on endangered species typical of heathland and acidic nutrient-poor grasslands. It is possible that soil variables other than NH4 mainly affect plant recruitment (De Graaf et al. 1997; Van den Berg, Vergeer & Roelofs 2003) whereas NH4 affects species at both the juvenile and adult stages. This would result in a gradual decline of populations exposed to, for example, high Al/Ca ratios and a rapid decline of populations exposed to high NH4 concentrations. This would explain why even adults of the rare species were absent from sites with high NH4 concentrations (Table 1). Other nutrient-poor habitat types in north-western Europe, such as mesotrophic marshes, soft-water lakes and ombrotrophic bogs, are also exposed to high NH4 deposition levels (Bobbink, Hornung & Roelofs 1998). From a conservation point of view it would be interesting to test whether the results of this study can be extrapolated to these habitats.

Our results highlight that conservation management in these nutrient-poor ecosystems needs to address the biogeochemical properties of the target sites. Standard conservation measures, such as tree and shrub clearance, restoration of the local hydrology and reinstatement of traditional grazing regimes, will fail to provide suitable habitat conditions for rare species in sites that have been exposed to long periods of elevated N deposition. Conservation management should therefore also aim to restore low NH4 concentrations and NH4/NO3 ratios. The species-specific mean levels of the NH4 concentration and NH4/NO3 ratio listed in Table 1 should probably be considered as the sustainable maximum that can be tolerated by species. It is likely that a number of samples in this study were from non-viable populations that would take the species-specific mean values we found for these variables to suboptimal levels. A range of experimental studies at various sites sampled in this study indicates that the most effective way to reduce both NH4 concentrations and NH4/NO3 ratios is to apply a combination of turf-cutting and liming (application of CaCO3; De Graaf et al. 1998b; Van den Berg, Vergeer & Roelofs 2003; Dorland et al. 2005). Turf-cutting removes a significant proportion of the accumulated NH4. Liming increases pH, which enhances soil nitrification rates, preventing new accumulation of NH4 and enhancing soil NH4/NO3 ratios. An additional benefit of liming is that it lowers soil Al/Ca ratios, thus removing a potential impediment to population persistence and growth of endangered species.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

We are most grateful to P. Goedhart for statistical advice and to R. Marrs and one anonymous referee for helpful comments on the manuscript. D. Kleijn was funded by the Stimulation Program Biodiversity.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
  • Aerts, R. & Berendse, F. (1988) The effects of increased nutrient availability on vegetation dynamics in wet heathlands. Vegetatio, 76, 6369.
  • Bekker, R.M. & Kwak, M.M. (2005) Life history traits as predictors of plant rarity, with particular reference to hemiparasitic Orobanchaceae. Folia Geobotanica, 40, 231242.
  • Bekker, R.M. & Lammerts, E.J. (2000) Naar een rode lijst met Groene stip voor hogere planten in Nederland. IKC-Natuurbeheer, Ministerie van Landbouw, Natuurbeheer en Visserij, Wageningen, the Netherlands.
  • Bobbink, R., Hornung, M. & Roelofs, J.G. (1998) The effects of air-borne nitrogen pollutants on species diversity in natural and semi-natural European vegetation. Journal of Ecology, 86, 717738.
  • Britto, D.T. & Kronzucker, H.J. (2002) inline image toxicity in higher plants: a critical review. Journal of Plant Physiology, 159, 567584.
  • Brown, J.H. (1984) On the relationship between abundance and distribution of species. American Naturalist, 124, 255279.
  • De Graaf, M.C.C., Bobbink, R., Roelofs, J.G.M. & Verbeek, P.J.M. (1998a) Differential effects of ammonium and nitrate on three heathland species. Plant Ecology, 135, 185196.
  • De Graaf, M.C.C., Bobbink, R., Verbeek, P.J.M. & Roelofs, J.G.M. (1997) Aluminium toxicity and tolerance in three heathland species. Water, Air and Soil Pollution, 98, 229239.
  • De Graaf, M.C.C., Verbeek, P.J.M., Bobbink, R. & Roelofs, J.G.M. (1998b) Restoration of species-rich dry heaths: the importance of appropriate soil conditions. Acta Botanica Neerlandica, 47, 89113.
  • De Graaf, M.C.C., Verbeek, P.J.M., Cals, M.J.R. & Roelofs, J.G.M. (1994) Effectgerichte maatregelen tegen verzuring en eutrofiëring van matig mineraalrijke heide en schraallanden. Vakgroep Oecologie, Werkgroep Milieubiologie, Katholieke Universiteit Nijmegen, Nijmegen, the Netherlands.
  • Dorland, E., Hart, M.A.C., Vermeer, M.L. & Bobbink, R. (2005) Assessing the success of wet heath restoration by combined sod cutting and liming. Applied Vegetation Science, 8, 209218.
  • Dueck, T. A. & Elderson, J. (1992) Influence of NH3 and SO2 on the growth and competitive ability of Arnica montana L. and Viola canina L. New Phytologist, 122, 507514.
  • Fennema, F. (1992) SO2 and NH3 deposition as possible causes for the extinction of Arnica montana L. Water, Air and Soil Pollution, 62, 325336.
  • Güsewell, S. & Koerselman, W. (2002) Variation in nitrogen and phosphorus concentrations of wetland plants. Perspectives in Plant Ecology, Evolution and Systematics, 5, 3761.
  • Houdijk, A.L.F.M., Verbeek, P.J.M., Van Dijk, H.F.G. & Roelofs, J.G.M. (1993) Distribution and decline of endangered herbaceous heathland species in relation to the chemical composition of the soil. Plant and Soil, 148, 137143.
  • Jansen, A.J.M., Eysink, F.T. & Maas, C. (2001) Hydrological processes in a CirsioMolinietum fen meadow: implications for restoration. Ecological Engineering, 17, 320.
  • Janssens, F., Peeters, A., Tallowin, J.R.B., Bakker, J.P., Bekker, R.M., Fillat, F. & Oomes, M.J.M. (1998) Relationship between soil chemical factors and grassland diversity. Plant and Soil, 202, 6978.
  • Kinraide, T.B. (2003) Toxicity effects in acidic forest soils: attempts to evaluate separately the toxic effects of excessive Al3+ and H+ and insufficient Ca2+ and Mg2+ upon root elongation. European Journal of Soil Science, 54, 323333.
  • Kunin, W.E. & Gaston, K.J. (1993) The biology of rarity: patterns, causes and consequences. Trends in Ecology and Evolution, 8, 298301.
  • Marrs, R.H., Phillips, J.D.P., Todd, P.A., Ghorbani, J. & Le Duc, M.G. (2004) Control of Molinia caerulea on upland moors. Journal of Applied Ecology, 41, 398411.
  • Matus, G., Verhagen, R., Bekker, R.M. & Grootjans, A.P. (2003) Restoration of the Cirsio dissectiMolinietum in the Netherlands: can we rely on soil seed banks? Applied Vegetation Science, 6, 7384.
  • Payne, R.W., Baird, D.B., Cherry, M., Gilmour, A.R., Harding, S.A., Kane, A.F., Lane, P.W., Murray, D.A., Soutar, D.M., Thompson, R., Todd, A.D., Tunnicliffe Wilson, G. & Welham, S.J. (2002) Genstat for Windows, 6th edn. VSN International, Oxford, UK.
  • Pearson, J. & Stewart, G.R. (1993) The deposition of atmospheric ammonia and its effects on plants. New Phytologist, 125, 283305.
  • Roelofs, J.G.M. (1986) The effect of airborne sulphur and nitrogen deposition on aquatic and terrestrial heathland vegetation. Experientia, 42, 372377.
  • Roelofs, J.G.M., Kempers, A.J., Houdijk, A.L.F.M. & Jansen, J. (1985) The effect of air-borne ammonium sulphate on Pinus nigra var. maritima in the Netherlands. Plant and Soil, 84, 4556.
  • Schaminée, J.H.J., Stortelder, A.H.F. & Weeda, E.J. (1996) De Vegetatie Van Nederland. 3. Plantengemeenschappen Van Graslanden, Zomen En Droge Heiden. Opulus Press, Leiden, the Netherlands.
  • Spek, T. (2005) Het Drentse esdorpenlandschap. Een historisch-geografische studie. PhD Thesis, Wageningen University, Wageningen, The Netherlands.
  • Strykstra, R.J., Pegtel, D.M. & Bergsma, A. (1998) Dispersal distance and achene quality of the rare anemochorous species Arnica montana L. Implications for conservation. Acta Botanica Neerlandica, 47, 4556.
  • Van Breemen, N. & Wright, R.F. (2004) History and prospect of catchment biogeochemistry: a European perspective based on acid rain. Ecology, 85, 23632368.
  • Van den Berg, L.J.L. (2006) Species-rich heathland degraded by atmospheric N deposition: perspectives for restoration. PhD thesis. Nijmegen University, Nijmegen, the Netherlands.
  • Van den Berg, L.J.L., Dorland, E., Vergeer, P., Hart, M.A.C., Bobbink, R. & Roelofs, J.G.M. (2005) Decline of acid-sensitive plant species in heathland can be attributed to ammonium toxicity in combination with low pH. New Phytologist, 166, 551564.
  • Van den Berg, L.J.L., Vergeer, P. & Roelofs, J.G.M. (2003) Heathland restoration in the Netherlands: effects of turf cutting depth on germination of Arnica montana. Applied Vegetation Science, 6, 117124.
  • Van der Meijden, R. (1990) Heukels’ Flora van Nederland. Wolters-Noordhoff, Groningen, the Netherlands.
  • Van der Meijden, R., Odé, B., Groen, C.L.G., Witte, J.-P.M. & Bal, D. (2000) Bedreigde en kwetsbare vaatplanten in Nederland. Basisrapport met voorstel voor de Rode Lijst. Gorteria, 26, 85208.
  • Vandvik, V., Heegaard, E., Måren, I.E. & Aarrestad, P.A. (2005) Managing heterogeneity: the importance of grazing and environmental variation on post-fire succession in heathlands. Journal of Applied Ecology, 42, 139149.
  • Vergeer, P., Regelink, R., Ouborg, N.J. & Roelofs, J.G.M. (2003) Effects of population size and genetic variation on the response of Succisa pratensis to eutrophication and acidification. Journal of Ecology, 91, 600609.
  • Walker, K.J., Pywell, R.F., Warman, E.A., Fowbert, J.A., Bhogal, A. & Chambers, B.J. (2004) The importance of former land use in determining successful re-creation of lowland heath in southern England. Biological Conservation, 116, 289303.
  • Wassen, M.J., Olde Venterink, H., Lapshina, E.D. & Tannenberger, F. (2005) Endangered plants persist under phosphorus limitation. Nature, 437, 547550.
  • Webb, N.R. (1998) The traditional management of European heathlands. Journal of Applied Ecology, 35, 987990.

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Table S1. The number of areas and samples used to calculate species’ means for a range of soil variables

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

FilenameFormatSizeDescription
JPE_1444_sm_AppendixS1.doc34KSupporting 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.