SEARCH

SEARCH BY CITATION

Keywords:

  • biotic interactions;
  • demography;
  • edge effects;
  • fitness;
  • population size

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  • 1
    Fragmentation reduces size and increases isolation of plant habitats, and increases the ratio between edge and centre area. Consequences of habitat fragmentation have rarely been studied for common plants, and edge effects are rarely studied for plants in general. We studied density, population structure, fitness components and biotic interactions in the locally abundant distylous fen plant Primula farinosa in the centres and at the edges of 27 Swiss fen habitats of different size and degree of isolation.
  • 2
    Population sizes ranged from 80 to 106 450 flowering plants and were larger in larger fen habitats than in smaller ones.
  • 3
    The densities of plants were lower in more isolated habitats (by 26–46% depending on developmental stage). In the less isolated habitats, the density of seedlings and juveniles was higher in larger habitats. Plant densities were lower at the edge than in the centre of habitats (34–55%), and edges had fewer plants of younger developmental states. Densities of reproductive plants differed only in the centres of habitats, where they were higher in larger habitats, while at the edges these densities were independent of habitat area.
  • 4
    Flower morph proportions were independent of size and isolation of habitats.
  • 5
    Plants in larger habitats had larger rosette diameters and tended to have more flowers.
  • 6
    At edges, seed set was on average 11% lower, and occurrence and degree of herbivory more than 50% higher, than in centres. Capsule frugivory was less likely in larger habitats. Infection by the smut fungus Urocystis primulicola was more likely in larger habitats.
  • 7
    We conclude that size, isolation and edge to centre ratio of fen habitats all affect the abundant P. farinosa and recommend that edge effects and more common species are given more attention.

Introduction

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

Habitat fragmentation reduces the size and increases the isolation of habitats and populations (Saunders et al. 1991). Stochasticity may affect small, isolated populations more strongly than larger ones (Shaffer 1987). Environmental stochasticity can drive small populations to extinction, while genetic stochasticity randomly reduces heterozygosity and increases the accumulation of deleterious mutations (Lande 1995; Lynch et al. 1995). Demographic stochasticity, i.e. random deviations in survival and reproduction of individuals from expected population means, may change sex ratios or proportions of style morphs in populations. Inbreeding, which can be pronounced in small, isolated populations, again reduces genetic variability and increases the accumulation of deleterious mutations. Furthermore, inbreeding depression can lower individual fitness and population viability (Ellstrand & Elam 1993; Young et al. 1996). Both biotic interactions (Olesen & Jain 1994), such as pollination, plant–herbivore, plant–pathogen and intraspecific interactions, and abiotic effects (Saunders et al. 1991) can be changed by habitat fragmentation in ways that may affect plant performance or population fitness, respectively. Differences in abiotic conditions between the edge and the centre of habitat remnants, affect larger proportions of smaller populations.

We studied several of these issues in the common wetland plant Primula farinosa L. (Primulaceae) in the centres and at the edges of 27 Swiss fen habitats of different size and degree of isolation. This clonal species is characteristic of calcareous fens (Caricion davallianae alliance; Ellenberg 1996). These are wetlands of high plant species diversity (c. 30 higher plant species per 2 m2; Pauli 1998) and among the few remaining seminatural ecosystems of Central Europe. In Switzerland, a 90% reduction of the wetland area since 1850 has caused high levels of destruction and fragmentation of habitats such as calcareous fens (Broggi & Schlegel 1989; Hintermann 1992). To date, habitat fragmentation has appeared to be more detrimental for rare, non-clonal, short-lived habitat specialists, than for perennial generalists with clonal reproduction (Fischer & Stöcklin 1997), such as P. farinosa. However, as the latter are major contributors to ecosystem productivity, any detrimental effects of fragmentation on such apparently well-buffered species would be cause for concern.

Demographic stochasticity causes the morph ratios of small and isolated populations of several heterostylous plants to deviate strongly from expected values (e.g. Barrett et al. 1989; Eckert & Barrett 1992; Ågren & Ericson 1996; Baker et al. 2000). Because plants of each morph are only compatible with plants of the other morph in distylous species, habitat fragmentation may limit mating opportunities and thus decrease reproduction (Barrett 1992). We studied whether deviations from the expected 1 : 1 morph ratio were more pronounced in smaller, more isolated populations of the distylous P. farinosa.

P. farinosa is subject to herbivory and frugivory and is host to the smut fungus Urocystis primulicola P. Magnus (Ustilaginaceae). Pollination may be lower in smaller, more isolated habitats because of lower pollinator diversity (Rathcke & Jules 1993; Ågren 1996) or a change in pollinator service (Heinrich 1979; Groom 1998; Murren 2002), thus reducing reproductive output. Conversely, because small habitats may harbour less diverse phytophagous insect communities (Bach 1988; Zabel & Tscharntke 1998), we expected less herbivore damage, for example by seed-predating insects (Jennersten & Nilsson 1993; Kéry et al. 2001). We also expected smut fungus to be less frequent in smaller populations (Jennersten et al. 1983; Burdon et al. 1995; Ericson et al. 1999; Groppe et al. 2001).

Whereas edge effects have long been recognized for animals and forest plants (e.g. Cadenasso et al. 1997; Lahti 2001), they have received little attention in studies on other plant species. We hypothesized that nutrient influx from surrounding agricultural land and altered pollination would result in larger plant size, but lower reproductive output and plant density, and more pronounced herbivory at habitat edges, but that differences between edges and centres would be less in smaller habitats.

Finally, we studied whether changes in plant size and density, reproductive output and pathogen incidence were reflected in lower fitness in more isolated, smaller fragments and at habitat edges. As adult plants are often less sensitive to changes than seedlings or juveniles (Oostermeijer et al. 1994; Jules 1998; Rose et al. 1998) and the effects of fragmentation may not therefore become apparent immediately (Eriksson 1996), it was necessary to consider the stage structure of P. farinosa (Saunders et al. 1991; Bühler & Schmid 2001).

We hypothesized that P. farinosa populations in smaller and more isolated habitats will show lower plant densities, altered morph and plant size structures, lower plant size, reproductive output and seed germination and lower leaf herbivory, frugivory and smut infection. We also predicted that edge effects would differ between large and small, and between more or less isolated populations.

Materials and methods

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

study species

Primula farinosa is widely but discontinuously distributed in the temperate and subarctic zone of Eurasia (Hegi 1906; Tutin et al. 1972). It is abundant in calcareous fens and wet meadows in the subalpine to alpine regions of Central Europe and in dry meadows in alpine regions (Hegi 1906). In Switzerland, P. farinosa is endangered at lower altitudes because of habitat destruction, but not at alpine altitudes (Landolt 1991). This herbaceous perennial with a basal rosette (Hegi 1906) may produce one to several daughter rosettes clonally. Leaves are subject to herbivory, mostly by snails and larvae of Lepidoptera (J. Lienert, personal observation). Plants of P. farinosa flower shortly after snow melt, typically in April or May. Flowering rosettes have one to several flower stalks 2–25 cm long, with up to 20 flowers in an umbel. Flowers are visited by various insects, including species of Coleoptera, Lepidoptera, Diptera and Hymenoptera (especially Bombus; J. Lienert, personal observation), many of them with long mouthparts.

While homostylous populations were found in Northern Europe (Mazer & Hultgard 1993), P. farinosa is generally distylous, in line with the heteromorphic self-incompatibility system usually found in Primula (Arnold & Richards 1998). So-called pin flowers are long-styled and have short anthers, whereas thrum flowers are short-styled with anthers positioned above the style. Fruit capsules contain up to 120 polyhedral seeds, which ripen within about 7–8 weeks. Fruit capsules are subject to frugivory by insect larvae (J. Lienert, personal observation).

Primula farinosa can be systemically infected by the anther smut Urocystis primulicola. Infected flowers have inconspicuous white conidia on anthers and the corolla tube (Nagler 1987). Infected fruit capsules are filled with powdery black spores and do not produce seeds (Nagler 1987; Pauli 1998). Pollinators could be responsible for dispersal of smut conidia (Nagler 1987), while the wind-distributed fungal spores are released from open capsules (Vánky 1994).

study sites and habitat characteristics

We studied 27 populations of P. farinosa in fens in two regions of north-east Switzerland: Ostschweiz (cantons AR, SG) and Innerschweiz (canton SZ; Table 1). All fens were of rectangular or circular shape and generally were bordered by nutrient-rich grassland containing nutrient-rich indicator species but no typical fen species (Lienert, Fischer & Diemer 2002). Populations of P. farinosa extended throughout each fen area. While the matrix surrounding fens was formerly used for low-intensity agriculture, today it mostly consists of intensively fertilized and drained agricultural land (Table 1). This can influence fen remnants through increased nutrient influx and altered groundwater levels. Traditionally, fens in Switzerland were mown once a year or extensively grazed by cattle and such management is now protected by governmental contracts. We restricted our study to fens that are mown annually, usually after September, and that are situated between 765 and 1250 m a.s.l.

Table 1.  Number of fen habitat, location (Swiss canton in parentheses: AR = Appenzell Ausserrhoden, SG = Sankt Gallen, SZ = Schwyz), name of fen, coordinates as in Swiss topographical maps, altitude, distance to nearest fen, degree of isolation (see methods), vegetation type surrounding fen or barrier type (‘pastures’ denotes sites that are intensively grazed by cattle), areal extent of population and population size (number of reproductive adults) for the 27 study sites of Primula farinosa
No.LocationName of fenLarge co-ordinateSmall co-ordinateAltitude (m asl)Distance (m)IsolationSurrounding vegetation type or barrierAreal extent (m2)Population size
1Wollerau (SZ)Unter Sennrüti 694 500225 825 770     600MoreAgricultural land/woodland   823  1 600
2Vorderthal (SZ)Chliweid710 425219 850 810> 1000MorePastures   853    850
3Euthal (SZ)Near Skilift704 925216 550 920    1000MoreAgricultural land 2 660  8 000
4Sattel (SZ)Gigersberg691 300213 5001010     200LessAgricultural land 3 500  5 500
5Sattel (SZ)In der Egg 1691 500214 0751020     210LessAgricultural land    50    135
6Sattel (SZ)In der Egg 2691 650214 1251030     210LessAgricultural land/woodland    40    100
7Gross (SZ)Breiten/Rotmoos700 750217 950 960> 1000MoreAgricultural land/woodland 8 750  4 000
8Einsiedeln (SZ)Eigenriet697 850215 900 990     700MoreAgricultural land16 633 19 000
9Willerzell (SZ)Sulzel703 400222 300 945     700MorePastures 7 000 13 300
10Alpthal (SZ)Sunnenberg697 475215 125 975     250LessAgricultural land /road/river 1 500     80
11Alpthal (SZ)Etteren696 975213 2501020> 1000MorePastures/agricultural land   525    170
12Hemberg (SG)Near Matt730 850239 800 940     800MoreAgricultural land/woodland/river 3 050  3 500
13Hemberg (SG)Near Brunnau731 600241 525 770> 1000MoreAgricultural land   126    150
14Urnäsch (AR) Röhrenmoos740 800242 200 935> 1000MorePastures/agricultural land/woodland14 400 12 100
15Krummenau (SG)Ämelsberg731 550235 000 930> 1000MorePastures/shrubs 3 700  3 000
16St. Peterzell (SG)Dreien, high731 775243 275 800     200LessAgricultural land   510    300
17St. Peterzell (SG)Dreien, low731 675243 400 765     200LessAgricultural land/road   660  1 810
18Vord. Laad (SG)Near farmhouse Egg733 850230 000 985     300LessPastures/agricultural land 1 600  4 500
19Ebnat Kappel (SG)Chellen/Allmeindswald730 650237 3501080     300LessWoodland16 500 94 500
20Sattel (SZ)Zäll692 050213 9751130     200LessAgricultural land/woodland19 000 40 600
21Schwyz (SZ)Haggen692 725212 1501220     300LessPastures   300  1 350
22Alpthal (SZ)Rund Blätz696 950210 9001210     400MoreWoodland19 000 64 600
23Grabs (SG)Maienberg748 300226 5751190     900MorePastures 2 030  4 100
24Wildhaus (SG)Riet/Lisighaus743 300229 4001010     300LessAgricultural land /shrubs/river 4 500 29 250
25Grabs (SG)Hagersriet, Voralpsee746 825225 4001250    1000MoreWoodland13 860 24 150
26Wildhaus (SG)Bilchenmoos748 100230 4001210     800MorePastures/woodland46 300106 450
27Urnäsch (SG)Below Egg, near Widleren735 900240 9501000     200LessPastures/woodland/river 6 450 17 200

Because of differences in topography and vegetation in the surrounding matrix, linear distance between fens was not an appropriate measure of habitat isolation. Hence, we classified 15 populations that were separated from any other population by more than 600 m of grassland or more than 300 m of wood- or shrubland as ‘more isolated’ (Table 1). The mean distance (1060 m ± SE 103 m) to the nearest fen for the 15 ‘more isolated’ populations was much larger than the mean distance (239 m ± 14 m) to the nearest fen for the 12 ‘less isolated’ ones.

To determine population size and mean population density (number of reproductive adults m−2), we counted all reproductive plants in small populations. In large populations, we counted the number of reproductive plants in 40–90 randomly selected 1 m2 plots and extrapolated to a value for the total fen area. The range of population sizes was similar in less (80–94 500 flowering plants) and more isolated populations (150–106 450 flowering plants; Table 1). Population sizes increased with increasing elevation (P < 0.05), reflecting the fact that habitat fragmentation was less severe at higher than lower altitude in Switzerland (Hintermann 1992; Bühler & Schmid 2001; Lienert, Fischer & Diemer 2002).

In May 2000, we selected 10 plants of P. farinosa in the centre and 10 at the edge of each population. Edge plants were those located nearest to 10 positions at equal distances along the perimeter. From each of the 10 edge plants we randomly threw a stick towards the centre of the fen and selected the plant closest to the stick as one of 10 centre plants. The distance between edge and centre plants was 5–15 m. We tagged the selected plants and collected fruits in June/July 2000. If plants consisted of several rosettes, we marked one randomly selected rosette.

population structure ofprimula farinosa

We distinguished three developmental states according to the following criteria (Gatsuk et al. 1980; Rabotnov 1985): (i) seedlings or juveniles with less than four primary leaves (i.e. subadults); (ii) vegetative adults with four or more leaves in a rosette; and (iii) reproductive adults with at least one flowering stem. As P. farinosa sometimes grows in clumps, we regarded all rosettes that were less than 1 cm apart as parts of the same clone. This was supported by the finding that they always produced the same flower morph and that rosette connections were sometimes visible. This was not the case for rosettes more than 1 cm apart.

To assess plant density of the different developmental states we counted the number of individuals in each state in 20 × 20 cm areas centred on our target plants in May 2000. As an additional measure of local density, we recorded the distance from each marked individual to the nearest conspecific neighbour.

fitness measures

We assessed the following vegetative and reproductive traits in May 2000: number of leaves per rosette, rosette diameter, length of the longest leaf, height of the flowering stem, number of rosettes per clone, number of flowers per inflorescence and flower morph type.

In late June and early July 2000, c. 8 weeks after the first measurements, we determined the fruit set of each marked plant in the field (fruit set = number of full capsules/(number of full + empty capsules)). We randomly collected one fruit per target plant, dried the capsules at room temperature for 8 weeks and counted the number of unfertilized ovules (very small), aborted seeds (large, polyhedral, brown) and developed seeds (large, polyhedral, orange; J. Lienert, personal observation). We used their sum to obtain the total number of ovules per capsule. We determined mean individual seed mass of developed seeds to the nearest µg. We also calculated fertilization rate ((number of aborted seeds + developed seeds)/total number of ovules), seed set (number of developed seeds/total number of ovules) and proportion of developed seeds (number of developed seeds/(number of aborted seeds + developed seeds)).

We collected seeds from all target plants for a germination experiment. Because capsules were lost due to frugivory in many populations and due to hail in population 17 (Table 1), we randomly selected further samples to obtain capsules of at least 20 plants per population. In late September 2000 we placed all seeds on wet Vermiculite (Vermex, Vermica AG, Bözen, Switzerland), with seeds from different capsules in different Petri dishes. We kept the seeds in the dark at +4 °C for 5 days, stratified them at −4 °C for 4 weeks, let them thaw at +4 °C for 5 days and placed them in a climate chamber under a 16-h light (20 °C)/8-h dark (10 °C) cycle. We randomized Petri dishes weekly. We recorded the proportion of seeds that had germinated at 1, 2, 3 and 4 weeks, after which hardly any additional seedlings appeared. We omitted 19 of 540 capsules from the analysis because of unripe seeds.

herbivory, frugivory and smut infection

From late June to early July 2000 we determined whether leaves and petals of target plants in the field had been subject to herbivory, and we estimated the degree of leaf herbivory as a percentage of leaf area lost. We also determined visually whether collected capsules had been subject to frugivory and assigned the extent to four classes: (0) no frugivory, (1) little frugivory, (2) moderate frugivory, and (3) pronounced frugivory. In the laboratory we checked again whether capsules showed traces of frugivory (such as holes, threads or insects) and assigned the degree of capsule frugivory to the same four classes. Moreover, we checked whether any plants (i.e. not only the 20 selected ones) in the populations showed signs of infection by the smut fungus U. primulicola.

data analysis

We tested relationships between habitat area, isolation and edge vs. centre, and plant density, plant size, reproduction, herbivory and pathogen infection using hierarchical analyses of variance (anova) and logistic regression (see skeleton anova, Table 2). Where necessary, we log-transformed data. We analysed proportions, binary and categorical data with logistic regression and treated mean deviance changes like mean squares in a normal anova, because their ratios approximately follow the F distribution (Payne et al. 1993). For binary data we used a complementary-log-log link model following Egli & Schmid (2001).

Table 2.  Skeleton hierarchical anova for effects on measures of density, plant size, reproduction, herbivory and pathogen infection of Primula farinosa
Sources of variationd.f.Variance ratios (F-values)
  1. For explanations on the sources of variation see methods (d.f. = degrees of freedom, MS = mean squares).

ln habitat area  1MShabitat area/MShabitat
Isolation  1MSisolation/MShabitat
ln habitat area × isolation interaction  1MShabitat area × isolation/MShabitat
Habitat 23MShabitat/MSresidual
Edge  1MSedge/MSedge × habitat
Edge × ln habitat area interaction  1MSedge × habitat area/MSedge × habitat
Edge × isolation interaction  1MSedge × isolation/MSedge × habitat
Edge × ln habitat area × isolation interaction  1MSedge × habitat area × isolation/MSedge × habitat
Edge × habitat interaction 23MShabitat/MSresidual
Residual540–54 

To account for differences between regions and elevation that may mask effects of habitat fragmentation, we considered region and elevation as covariates. We conducted all analyses with and without covariates, and tested the effects of covariates against the remaining variation between habitats. Because results did not change, we present the results without covariates. We used the computer program GenStat 5, release 3.2. (Payne et al. 1993).

Results

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

population size and density

The number of flowering plants of P. farinosa was higher in larger habitats (P < 0.0001, Table 3, Fig. 1a), while plant density was related to both isolation and size of habitats. The mean density of plants of all developmental states was lower in more isolated habitats: the mean density of reproductive adults over the whole site was lower by 46% (P < 0.05, Fig. 1b), mean density of vegetative adults was lower by 34% (P < 0.05, Fig. 1c) and mean total density was lower by 26% (P < 0.01, Fig. 1d). The mean density of subadults was greater in larger habitats that were less isolated (habitat area × isolation interaction, P < 0.05), while in more isolated habitats there was no relationship between the density of subadults and habitat area (Fig. 1e).

Table 3.  Effects of habitat area, isolation, edge and interactions between these factors on density measures of Primula farinosa in 27 populations. The observed traits are: population size (number of reproductive adults), density over the whole site (counted as reproductive adults m−2), total density (i.e. of all developmental states), density of only reproductive adults, density of only vegetative adults, density of only subadults in 20 sample quadrats of 20 × 20 cm per population, proportion of flowering adults, proportion of vegetative adults, proportion of subadults to the total number of plants, proportion of reproductive adults to the total number of adults and distance of the marked plant to the nearest plant of P. farinosa (cm; see methods)
Sources of variationd.f.ln population size MSln density whole site (reproductive adults) MSln total density (all developmental states) MSln density reproductive adults MSln density vegetative adults MSln density sub-adults MSReproductive adults/all developmental states MDVegetative adults/all developmental states MDSub-adults/all developmental states MDReproductive adults/ all adults MDln distance to next Primula MS
  • a

    d.f. ln distance to next Primula= 485. All density measures except proportions were ln-transformed. For error terms see Table 2. We present mean squares (MS) or mean deviance changes (MD) and P-values:

  • (*)

    P < 0.1;

  • *

    P < 0.05;

  • **

    P < 0.01;

  • ***

    P < 0.001.

ln Habitat area  196.32***0.2110.020.0000.010.460.000.110.160.06 5.81
Isolation  1 1.390.994*4.51**0.511(*)4.18*1.51(*)1.540.740.451.84 3.36
ln habitat area × isolation  1 0.100.0111.440.0561.842.48*4.35(*)1.292.563.44(*) 0.24
Habitat 23 1.030.2130.55***0.1550.94***0.44**1.30*0.760.88***1.11 3.22***
Edge  19.32***0.2168.23***4.31***11.06***3.85*5.11**8.45***64.21***
Edge × ln habitat area  10.650.958*0.010.720.190.800.620.46 6.74(*)
Edge × isolation  11.080.0041.80*0.350.570.640.101.72(*) 3.23
Edge × ln hab. area × isolation  10.360.2610.080.430.450.990.080.48 0.05
Edge × habitat 230.380.2110.380.32(*)0.650.480.58**0.51 1.97*
Residual486a0.220.1650.300.220.790.690.290.82 1.20
image

Figure 1. Population size and density measures for 27 populations of Primula farinosa in north-east Switzerland in relation to (a) fen habitat area, (b–d) isolation of populations and (e) their interaction. (a) Population size measured as number of flowering plants. (b) Density of reproductive adults over the whole site. (c) Density of vegetative adults. (d) Total density (i.e. of plants of all developmental states). (e) Density of subadults. In b–d means ±se are shown. Filled symbols in e denote less isolated habitats.

Download figure to PowerPoint

Mean density of P. farinosa was lower at habitat edges by 34% for total density (P < 0.001, Table 3, Fig. 2a), by 40% for vegetative adults (P < 0.001) and by 55% for subadults (P < 0.001, Fig. 2b). The distance to the nearest conspecific neighbour was 73% lower in the centre than at the edge (P < 0.001). The proportion of reproductive plants was 22% lower in the centre than at the edge (P < 0.001), where proportions of vegetative adults were 21% lower (P < 0.05) and of subadults 43% lower (P < 0.01, Fig. 2c) than in the centre. Proportions of reproductive adults to the total number of adults were 18% lower in the centre (P < 0.001). The lower proportions of earlier developmental states at edges were due to lower absolute numbers of subadult and vegetative plants, whereas absolute numbers of flowering adults did not differ between edge and centre habitats. In the centre, but not at the edge, density of reproductive adults increased with increasing habitat area (edge × habitat area interaction, P < 0.05, Table 3, Fig. 2d). At the edge of less isolated habitats, the density of vegetative adults was 49% lower than in the centre, but in more isolated habitats it was only 27% lower at the edge (isolation × edge interaction, P < 0.05, Fig. 2e).

image

Figure 2. Density measures of 27 populations of Primula farinosa in relation to the position of individual plants in the centre or at the edge of habitats. (a) Total density (i.e. all developmental states). (b) Density of subadults. (c) Proportion of subadults to the total number of plants. (d) Density of reproductive adults also in relation to habitat area. (e) Density of vegetative adults also in relation to habitat isolation (± SE). Shaded bars and filled symbols denote centres of habitats.

Download figure to PowerPoint

plant size and morph

Plants in larger habitats had larger rosette diameters (P < 0.05, Table 4, Fig. 3a) and tended to have more flowers and fewer but longer leaves (all P < 0.1). Stem lengths increased more strongly with increasing habitat area at the edge than in the centre of habitats (edge × habitat area interaction, P < 0.05). In less isolated habitats the number of leaves per plant (edge × isolation interaction, P < 0.05), rosette diameter (P < 0.05), stem height (P < 0.001) and the number of flowers (P < 0.01, Fig. 3b) were larger at the edge than in the centre, whereas in more isolated habitats they were smaller at the edge than in the centre. Proportions of pin plants, which ranged from 14.3% to 73.3% per population, did not differ significantly between populations and between centre and edge (Table 4).

Table 4.  Effects of habitat area, isolation, edge and interactions between these factors on field measures of plant size and reproduction of Primula farinosa in 27 populations. The observed traits are: number of leaves per rosette, ln-transformed rosette diameter (cm), leaf length (cm), stem height (cm), number of rosettes per clone, ln-transformed number of flowers per main stem, flower morph (pin or thrum) and fruit set, measured as the ratio (full capsules/(full + empty capsules))
Sources of variationd.f.Number of leaves MSln rosette diameter MSLeaf length MSStem height MSNumber of rosettes MDln number of flowers MSFlower morph MDFruit set MD
  • a

    d.f. Number of leaves = 485; d.f. stem height, flower morph = 326; d.f. number of flowers = 329; d.f. fruit set = 281. We present mean squares (MS) or mean deviance changes (MD) and P-values:

  • (*)

    P < 0.1;

  • *

    P < 0.05;

  • **

    P < 0.01;

  • ***

    P < 0.001.

ln habitat area  1199.0(*)1.48*11.09(*)129.90.8071.779(*)1.0470.005
Isolation  1 28.70.09 0.28 17.60.8820.0011.9260.001
ln habitat area × isolation  1  7.90.04 1.09145.80.6940.7970.1610.001
Habitat 23 52.1***0.269*** 3.70***117.1***0.342*0.473***1.3271.387***
Edge  1  7.90.001 0.72 17.8(*)0.0190.0271.5450.455
Edge × ln habitat area  1  0.10.126(*) 0.82 19.4*0.0020.0000.0451.333
Edge × isolation  1 48.8*0.193* 1.95(*) 68.0***0.0021.207**0.1920.003
Edge × ln habitat area × isolation  1  6.50.017 0.08  5.870.0290.0000.0910.040
Edge × habitat 23  6.90.041 0.60  4.220.2230.1441.2140.564
Residual486a  8.50.043 0.56  6.90.1950.1361.4100.560
image

Figure 3. Vegetative and reproductive fitness measures of plants of 27 populations of Primula farinosa in relation to area and degree of isolation of habitats and to the position of individual plants in the centre or at the edge of habitats: (a) rosette diameter; (b) number of flowers per rosette; and (c) seed set. In b and c means ±se are shown. Shaded bars denote habitat centres.

Download figure to PowerPoint

seed production and germination

Mean seed set was 11% lower at the edge of habitats (P < 0.05, Table 5, Fig. 3c). Patterns of germination percentage were the same after 1, 2, 3 and 4 weeks, except that after 1 week the germination percentage was 17% lower for capsules from the centre than from the edge of habitats, but not thereafter (P < 0.01). Germination percentage increased with habitat area in the centres of less isolated habitats, but not at edges or in centres of more isolated habitats (three-way interaction, P < 0.05, Table 5).

Table 5.  Effects of habitat area, isolation, edge and interactions between these factors on seed production and germination of Primula farinosa in 27 populations. The degrees of freedom for the edge × population interaction is 22, as nearly all capsules were damaged by hail in population 17 (St Peterzell, Dreien low). The observed traits are: number of ovules per capsule, number of developed seeds per capsule, ln-transformed number of aborted seeds per capsule, ln-transformed seed mass (µg), fertilization rate ((number of aborted seeds + developed seeds)/number of ovules), seed set (number of developed seeds/number of ovules), proportion of developed seeds (number of developed seeds/(number of aborted seeds + developed seeds)) and germination percentage after 4 weeks
Sources of variationd.f.No. of ovules MSNo. of developed seeds MSln no. of aborted seeds MSln seed mass MSFertilization rate MDSeed set MDProportion of developed seeds MDGermination percentage MD
  • a

    d.f. seed mass = 221; d.f. fertilization rate, seed set, proportion of developed seeds = 235; d.f. germination percentage = 214. We present mean squares (MS) or mean deviance changes (MD) and P-values:

  • (*)

    P < 0.1;

  • *

    P < 0.05;

  • **

    P < 0.01;

  • ***

    P < 0.001.

ln habitat area  1  1312980.0360.0360.7550.7900.2411.394
Isolation  1  82 2580.5550.0630.1770.3860.4510.119
ln habitat area × isolation  11717 4700.8870.0000.3670.5882.995*1.342
Habitat 232138***1442***0.7880.194(*)0.605**0.700***0.631***1.034***
Edge  1 392 8760.2350.0170.729(*)1.241*1.2870.717(*)
Edge × ln habitat area  1  24 4940.0150.0590.3820.4070.0060.639(*)
Edge × isolation  1 892(*) 1161.724(*)0.0000.0630.4120.8510.547(*)
Edge × ln habitat area × isolation  1 172 5020.0010.0240.1690.0950.1081.565*
Edge × habitat 22 291 4580.4910.1550.2450.2920.509**0.212
Residual236a 431 4870.6280.1350.3210.2900.2390.314

herbivory, frugivory and smut infection

Capsule frugivory was less pronounced in larger habitats (P < 0.05, Table 6, Fig. 4a). In the centre of habitats the occurrence of leaf herbivory was 35% lower (P < 0.001, Table 6), the degree of leaf herbivory was 55% lower (P < 0.001) and the occurrence of petal herbivory was 56% lower (P < 0.05) than at the edges. Moreover, while leaf herbivory at the edges of habitats was equally likely for all habitats, it was less likely in the centres of larger habitats (edge × habitat area interaction, P < 0.01, Fig. 4b). The decrease in occurrence and degree of leaf herbivory from edge to centre was more pronounced for less isolated than for more isolated habitats (edge × isolation interaction, both P < 0.05, Fig. 4c,d).

Table 6.  Effects of habitat area, isolation, edge and interactions between these factors on herbivory, frugivory and smut occurrence in 27 populations of Primula farinosa. The observed traits are: occurrence of leaf herbivory (yes/no), ln-transformed degree of leaf herbivory in percentage, occurrence of petal herbivory, degree of capsule frugivory in the field (in four classes; 0 = none; 1 = little; 2 = moderate; 3 = pronounced frugivory), degree of capsule frugivory in the laboratory (same classification) and occurrence of smut fungus in population
Sources of variationd.f.Occurrence leaf herbivory MDln degree leaf herbivory MSOccurrence petal herbivory MDDegree capsule frugivory, field MDDegree capsule frugivory, laboratory MDOccurence smut in population MD
  • a

    d.f. petal herbivory = 326; d.f. degree capsule frugivory in field = 281; d.f. degree capsule frugivory in laboratory = 237. We present mean squares (MS) or mean deviance changes (MD) and P-values:

  • (*)

    P < 0.1;

  • *

    P < 0.05;

  • **

    P < 0.01;

  • ***

    P < 0.001.

ln habitat area  1 9.1622.70(*)1.101.19 6.74*170.9**
Isolation  1 0.00 0.333.101.01 1.34 11.3
ln habitat area × isolation  1 0.13 0.700.961.68 0.00  0.2
Habitat 23 3.19*** 5.60***1.27***2.84(*) 1.41*** 20.7
Edge  147.95***71.13***6.59*1.31 1.62(*)
Edge × ln habitat area  1 8.31** 3.56(*)0.131.97 1.05
Edge × isolation  1 4.21* 5.81*3.20(*)0.27 1.05
Edge × ln habitat area × isolation  1 0.09 0.471.231.4010.12***
Edge × habitat 23 0.79 1.050.85**3.96** 0.50*
Residual475a 1.12 1.180.461.85 0.30
image

Figure 4. Damage by herbivores or frugivores and occurrence of pathogens in 27 populations of Primula farinosa in relation to (a, b, e) habitat area (b, c, d) position of individual plants in the centre or at the edge of habitats and (c, d) degree of habitat isolation. (a) Population means of the degree of frugivory of individual capsules (as classified in four classes in the laboratory: 0, no frugivory; 1, little frugivory; 2, moderate traces of frugivory; 3, pronounced frugivory. All population means were < 0.4). (b, c) Occurrence of leaf herbivory. (d) Degree of leaf herbivory in percentage of leaf area loss. (e) Occurrence of the smut fungus Urocystis primulicola in populations. Shaded bars and filled symbols denote centres of habitats.

Download figure to PowerPoint

Populations in larger habitats were more likely to host populations of the smut fungus Urocystis primulicola (P < 0.01, Table 6) and the smut fungus was only found in populations with more than 4100 reproductive adults (Fig. 4e). However, we did not find infected plants in five populations with 13 300–40 600 reproductive adults.

Discussion

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

population size, density and structure, and fitness measures

Larger fen habitats carried larger populations of P. farinosa (Table 3, Fig. 1a) and more isolated populations were less dense than less isolated populations (Fig. 1b–d). This suggests reduced rates of establishment, survival or reproduction in more isolated habitats of this abundant plant. Moreover, the density of seedlings and juveniles increased strongly with increasing habitat area in less isolated populations (Fig. 1e).

The smaller plant diameter and the tendency to have less flowers of plants in smaller habitats (Table 4, Fig. 3a) is in line with reports of reduced plant performance in small compared with large populations (Menges 1991; Aizen & Feinsinger 1994; Ouborg & van Treuren 1994; Fischer & Matthies 1998; Kéry et al. 2000). An explanation for the lower plant densities in more isolated populations of P. farinosa, despite similar adult fitness traits across more and less isolated habitats, could be lower seedling establishment. This may be caused by the combined effects of inbreeding depression and lower genetic diversity in more isolated populations (Ellstrand & Elam 1993). Alternatively, seedling establishment could be influenced by changed habitat characteristics in more isolated fragments. An ongoing common environment study with offspring of the plants studied here will allow us to separate genetic effects of habitat fragmentation on plant performance of P. farinosa from environmental ones that are possibly confounded with fragmentation.

In contrast to reports of strong deviations of morph ratios from expected values, especially in small populations of heterostylous species (Barrett et al. 1989; Eckert & Barrett 1992; Ågren & Ericson 1996; Barrett & Husband 1997; Mal & Lovett Doust 1997; Baker et al. 2000; Kéry 2000), flower morph ratios of P. farinosa did not deviate strongly from the expected 1 : 1 ratio and deviations were independent of habitat size and isolation. Possibly, our study populations, which all contained at least 80 reproductive adults (Table 1), were too large for a strong effect of demographic stochasticity on the morph ratio.

edge effects

We found strong edge effects in P. farinosa populations. Plant density, the proportion of plants in more juvenile developmental states (Table 3, Fig. 2) and seed set were all lower at edges than in habitat centres (Table 5, Fig. 3c), while herbivory was higher (Table 6, Fig. 4c,d). Lower proportions of plants in younger developmental states were due to lower absolute numbers of subadult and vegetative stages, which may jeopardize long-term population viability. Such a ‘regressive’ population structure was reported for fragmented populations of the rare Gentiana pneumonanthe (Oostermeijer et al. 1994; Rose et al. 1998) and the common Trillium ovatum (Jules 1998), Carex davalliana and Succisa pratensis (Hooftman & Diemer 2002), but not for the locally abundant fen specialist Swertia perennis (Lienert, Diemer & Schmid 2002). For these species, however, the role of edge effects is unknown. Longer-term demographic studies relating population growth rates to individual fitness components are needed to determine whether reduced densities of subadults indicate reduced population viability. The ongoing common environment study with P. farinosa will allow us to separate possible effects of genetic differentiation between edge and centre plants from purely environmental effects. Environmental edge effects, which may play a role, could be due to nutrient influx from surrounding agricultural land into habitat edges.

Increasing habitat area was associated with higher density of reproductive plants and decreasing occurrence of leaf herbivory in the centre, but not at the edge of habitats (Figs 2d and 4b). The smallest P. farinosa habitats in today's fragmented landscape may consist, to a large degree, of ‘edge’.

biotic interactions in fragmented populations

Capsule frugivory was lower in larger habitats of P. farinosa (Table 6, Fig. 4a), contradicting results from studies with Viscaria vulgaris (Jennersten & Nilsson 1993) and Gentiana cruciata (Kéry et al. 2001). In our populations, the higher plant-species richness of fen remnants may attract herbivores from the surrounding habitat, and the smaller area relative to the matrix may lead to higher herbivore densities in smaller fen remnants. Larger sites may also support more predators of herbivores.

A lower degree of herbivory in more isolated populations was found in populations of Vicia sepium (Kruess & Tscharntke 2000) and Clarkia concinna concinna (Groom 2001), but the degree of habitat isolation did not affect herbivory in P. farinosa. However, our study populations contained at least 80, and usually many more than 100, plants (Table 1), whereas the previous study populations had < 25 individuals. Nevertheless, relatively higher leaf herbivory at the edges of less isolated habitats (edge × isolation interaction, Figs 4c, d) suggests that a lower degree of habitat isolation increases herbivore attack on P. farinosa.

As in P. farinosa (Fig. 4e), stronger infection by smut fungi in larger or more dense populations has been observed in Viscaria vulgaris (Jennersten et al. 1983), Valeriana sambucifolia (Carlsson et al. 1990), Silene alba (Antonovics et al. 1994), Filipendula ulmaria (Burdon et al. 1995) and Valeriana salina (Ericson et al. 1999). Pronounced stochastic fluctuations in small fungal populations may increase their extinction rates and explain the lower likelihood of pathogen infection in small plant populations (Ericson et al. 1999). Alternatively, larger plant populations might attract more foraging insects carrying fungal spores (Jennersten et al. 1983). While smaller populations of P. farinosa were less likely to host the smut fungus than larger populations, plants in smaller populations might be more susceptible to smut infection because of higher degrees of inbreeding (Ouborg et al. 2000).

Conclusions

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

Edge effects and effects of small size and isolation of Primula farinosa habitats on plant and population performance indicate that habitat fragmentation affects even this locally abundant species. In particular the effects on biotic interactions, which have consequences for plant fitness, suggest that such interactions should not be neglected in studies of plant responses to habitat fragmentation. Our results imply that there will be further changes in plant and population performance with further habitat fragmentation. We recommend that future studies on habitat fragmentation and conservation activities consider edge effects and do not only focus on rare species.

Acknowledgements

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

The authors thank Andrea Lienert and Judith Vonwil for patiently counting seeds and seedlings and Lindsay Haddon, Michael Hutchings and two anonymous referees for helpful comments on an earlier version of the manuscript. We also thank farmers, landowners, nature conservancy agencies and municipal authorities for allowing us to work on their land. This study was financed by the Swiss National Science Foundation (grant 31–56809.99).

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  • Ågren, J. (1996) Population size, pollinator limitation, and seed set in the self-incompatible herb Lythrum salicaria. Ecology, 77, 17791790.
  • Ågren, J. & Ericson, L. (1996) Population structure and morph-specific differences in tristylous Lythrum salicaria. Evolution, 50, 126139.
  • Aizen, M.A. & Feinsinger, P. (1994) Forest fragmentation, pollination, and plant reproduction in a chaco dry forest, Argentina. Ecology, 75, 330351.
  • Antonovics, J., Thrall, P.H., Jarosz, A.M. & Stratton, D. (1994) Ecological genetics of metapopulations: the Silene-Ustilago plant-pathogen system. Ecological Genetics (ed. L.A.Real), pp. 146170. Princeton University Press, Princeton, New Jersey.
  • Arnold, E.S. & Richards, A.J. (1998) On the occurrence of unilateral incompatibility in Primula section Aleuritia Duby and the origin of Primula scotica Hook. Botanical Journal of the Linnean Society, 128, 359368.
  • Bach, C.E. (1988) Effects of host plant patch size on herbivore density: patterns. Ecology, 69, 10901102.
  • Baker, A.M., Thompson, J.D. & Barrett, S.C.H. (2000) Evolution and maintenance of stigma-height dimorphism in Narcissus. I. Floral variation and style-morph ratios. Heredity, 84, 502513.
  • Barrett, S.C.H. (1992) Heterostylous genetic polymorphisms: model systems for evolutionary analysis. Evolution and Function of Heterostyly (ed. S.C.H.Barrett), pp. 129. Springer, Berlin.
  • Barrett, S.C.H. & Husband, B.C. (1997) Ecology and genetics of ephemeral plant populations: Eichhornia paniculata (Pontederiaceae) in northeast Brazil. Journal of Heredity, 88, 277284.
  • Barrett, S.C.H., Morgan, M.T. & Husband, B.C. (1989) The dissolution of a complex genetic polymorphism: the evolution of self-fertilization in tristylous Eichhornia paniculata (Pontederiaceae). Evolution, 43, 13981416.
  • Broggi, M.F. & Schlegel, H. (1989) Mindestbedarf an Naturnahen Flächen in der Kulturlandschaft. Bericht 31 des Nationalen Forschungsprogrammes ‘Nutzung des Bodens in der Schweiz’, Liebefeld, Bern.
  • Bühler, C. & Schmid, B. (2001) The influence of management regime and altitude on the population structure of Succisa pratensis: implications for vegetation monitoring. Journal of Applied Ecology, 38, 689698.
  • Burdon, J.J., Ericson, L. & Müller, W.J. (1995) Temporal and spatial changes in a metapopulation of the rust pathogen Triphragmium ulmariae and its host, Filipendula ulmaria. Journal of Ecology, 83, 979989.
  • Cadenasso, M.L., Traynor, M.M. & Pickett, S.T.A. (1997) Functional location of forest edges: gradients of multiple physical factors. Canadian Journal of Forest Research, 27, 774782.
  • Carlsson, U., Elmqvist, T., Wennström, A. & Ericson, L. (1990) Infection by pathogens and population age of host plants. Journal of Ecology, 78, 10941105.
  • Eckert, C.G. & Barrett, S.C.H. (1992) Stochastic loss of style morphs from populations of tristylous Lythrum salicaria and Decodon verticillatus (Lythraceae). Evolution, 46, 10141029.
  • Egli, P. & Schmid, B. (2001) The analysis of complex leaf survival data. Basic and Applied Ecology, 2, 223231.
  • Ellenberg, H. (1996) Vegetation Mitteleuropas Mit Den Alpen in Ökologischer, Dynamischer und Historischer Sicht. Ulmer, Stuttgart.
  • Ellstrand, N.C. & Elam, D.R. (1993) Population genetic consequences of small population size: implications for plant conservation. Annual Review of Ecology and Systematics, 24, 217242.
  • Ericson, L., Burdon, J.J. & Müller, W.J. (1999) Spatial and temporal dynamics of epidemics of the rust fungus Uromyces valerianae on populations of its host Valeriana salina. Journal of Ecology, 87, 649658.
  • Eriksson, O. (1996) Regional dynamics of plants: a review of evidence for remnant, source-sink and metapopulations. Oikos, 77, 248258.
  • Fischer, M. & Matthies, D. (1998) Effects of population size on performance in the rare plant Gentianella germanica. Journal of Ecology, 86, 195204.
  • Fischer, M. & Stöcklin, J. (1997) Local extinctions of plants in remnants of extensively used calcareous grasslands 1950–85. Conservation Biology, 11, 727737.
  • Gatsuk, L.E., Smirnova, O.V., Vorontzova, L.I., Zaugolnova, L.B. & Zhukova, L.A. (1980) Age states of plants of various growth forms: a review. Journal of Ecology, 68, 675696.
  • Groom, M.J. (1998) Allee effects limit population viability of an annual plant. American Naturalist, 151, 487496.
  • Groom, M.J. (2001) Consequences of subpopulation isolation for pollination, herbivory, and population growth in Clarkia concinna concinna (Onagraceae). Biological Conservation, 100, 5563.
  • Groppe, K., Steinger, T., Schmid, B., Baur, B. & Boller, T. (2001) Effects of habitat fragmentation on choke disease (Epichloë bromicola) in the grass Bromus erectus. Journal of Ecology, 89, 247255.
  • Hegi, G. (1906) Primula. Illustrierte Flora Von Mitteleuropa, Vol. 5, pp. 17331758. J.F. Lehmanns Verlag, München.
  • Heinrich, B. (1979) Resource heterogeneity and patterns of movement in foraging bumblebees. Oecologia, 40, 235245.
  • Hintermann, U. (1992) Schlussbericht zum Inventar der Moorlandschaften von Besonderer Schönheit und von Nationaler Bedeutung. Bundesamt für Umwelt, Wald und Landschaft (BUWAL), Schriftenreihe Umwelt 168, Bern.
  • Hooftman, D.A.P. & Diemer, M. (2002) Effects of small habitat size and isolation on the population structure of common wetland species. Plant Biology, 4, 720728.
  • Jennersten, O. & Nilsson, S.G. (1993) Insect flower visitation frequency and seed production in relation to patch size of Viscaria vulgaris (Caryophyllaceae). Oikos, 68, 283292.
  • Jennersten, O., Nilsson, S.G. & Wästljung, U. (1983) Local plant populations as ecological islands: the infection of Viscaria vulgaris by the fungus Ustilago violacea. Oikos, 41, 391395.
  • Jules, E.S. (1998) Habitat fragmentation and demographic change for a common plant: Trillium in old-growth forest. Ecology, 79, 16451656.
  • Kéry, M. (2000) Ecology of small populations. PhD thesis, Universität Zürich, Zürich.
  • Kéry, M., Matthies, D. & Fischer, M. (2001) The effect of plant population size on the interactions between the rare plant Gentiana cruciata and its specialized herbivore Maculinea rebeli. Journal of Ecology, 89, 418427.
  • Kéry, M., Matthies, D. & Spillman, H.H. (2000) Reduced fecundity and offspring performance in small populations of the declining grassland plants Primula veris and Gentiana lutea. Journal of Ecology, 88, 1730.
  • Kruess, A. & Tscharntke, T. (2000) Species richness and parasitism in a fragmented landscape: experiments and field studies with insects on Vicia sepium. Oecologia, 122, 129137.
  • Lahti, D.C. (2001) The ‘edge effect on nest predation hypothesis’ after twenty years. Biological Conservation, 99, 365374.
  • Lande, R. (1995) Mutation and conservation. Conservation Biology, 9, 782791.
  • Landolt, E. (1991) Gefährdung der Farn- und Blütenpflanzen in der Schweiz mit Gesamtschweizerischen und Regionalen Roten Listen. Bundesamt für Umwelt, Wald und Landschaft (BUWAL), Eidgenössische Drucksachen- und Materialzentrale (EDMZ), Bern.
  • Lienert, J., Diemer, M. & Schmid, B. (2002) Effects of habitat fragmentation on population structure and fitness components of the wetland specialist Swertia perennis L. (Gentianaceae). Basic and Applied Ecology, 3, 101114.
  • Lienert, J., Fischer, M. & Diemer, M. (2002) Local extinctions of the wetland specialist Swertia perennis L. (Gentianaceae) in Switzerland: a revisitation study based on herbarium records. Biological Conservation, 103, 6576.
  • Lynch, M., Conery, J. & Bürger, R. (1995) Mutation accumulation and the extinction of small populations. American Naturalist, 146, 489518.
  • Mal, T.K. & Lovett Doust, J. (1997) Morph frequencies and floral variation in a heterostylous colonizing weed, Lythrum salicaria. Canadian Journal of Botany, 75, 10341045.
  • Mazer, S.J. & Hultgard, U.-M. (1993) Variation and covariation among floral traits within and among four species of Northern European Primula (Primulaceae). American Journal of Botany, 80, 474485.
  • Menges, E.S. (1991) Seed germination percentage increases with population size in a fragmented prairie species. Conservation Biology, 5, 158164.
  • Murren, C.J. (2002) Effects of habitat fragmentation on pollination: pollinators, pollinia viability and reproductive success. Journal of Ecology, 90, 100107.
  • Nagler, A. (1987) Urocystis Rabenhorst und Ginanniella Ciferri zwei eigenständige Gattungen? Urocystis galanthi Pape und Ginanniella primulae (Rostrup) Ciferri. Zeitschrift für Mykologie, 53, 331354.
  • Olesen, J.M. & Jain, S.K. (1994) Fragmented plant populations and their lost interactions. Conservation Genetics (eds V.Loeschcke, J.Tomiuk & S.K. Jain), pp. 417426. Birkhäuser, Basel.
  • Oostermeijer, J.G.B., Van’t Veer, R. & Den Nijs, J.C.M. (1994) Population structure of the rare, long-lived perennial Gentiana pneumonanthe: relation to vegetation and management in the Netherlands. Journal of Applied Ecology, 31, 428438.
  • Ouborg, N.J., Biere, A. & Mudde, C.L. (2000) Inbreeding effects on resistance and transmission-related traits in the Silene-Microbotryum pathosystem. Ecology, 81, 520531.
  • Ouborg, N.J. & Van Treuren, R. (1994) The significance of genetic erosion in the process of extinction. IV. Inbreeding load and heterosis in relation to population size in the mint Salvia pratensis. Evolution, 48, 9961008.
  • Pauli, D. (1998) Plant species diversity and productivity in wetland communities: patterns and processes. PhD thesis, Universität Zürich, Zürich.
  • Payne, R.W., Lane, P.W., Digby, P.G.N., Harding, S.A., Leech, P.K., Morgan, G.W. et al. (1993) Genstat 5 Reference Manual, Release 3. Oxford Science Publications, Clarendon Press, Oxford.
  • Rabotnov, T.A. (1985) Dynamics of plant coenotic populations. The Population Structure of Vegetation (ed. J.White), pp. 121142. Dr W. Junk Publishers, Dordrecht.
  • Rathcke, B.J. & Jules, E.S. (1993) Habitat fragmentation and plant–pollinator interactions. Current Science, 65, 273277.
  • Rose, R.J., Clarke, R.T. & Chapman, S.B. (1998) Individual variation and the effects of weather, age and flowering history on survival and flowering of the long-lived perennial Gentiana pneumonanthe. Ecography, 21, 317326.
  • Saunders, D.A., Hobbs, R.J. & Margules, C.R. (1991) Biological consequences of ecosystem fragmentation: a review. Conservation Biology, 5, 1832.
  • Shaffer, M. (1987) Minimum viable populations: coping with uncertainty. Viable Populations for Conservation (ed. M.E.Soulé), pp. 6986. Cambridge University Press, Cambridge.
  • Tutin, T.G., Heywood, V.H., Burges, N.A., Moore, D.M., Valentine, D.H., Walters, S.M et al. (1972) Flora Europaea, Vol. 3. Cambridge University Press, Cambridge.
  • Vánky, K. (1994) European Smut Fungi. Gustav Fischer Verlag, Stuttgart.
  • Young, A.G., Boyle, T. & Brown, T. (1996) The population genetic consequences of habitat fragmentation for plants. Trends in Ecology and Evolution, 11, 413418.
  • Zabel, J. & Tscharntke, T. (1998) Does fragmentation of Urtica habitats affect phytophagous and predatory insects differentially? Oecologia, 116, 419425.