Make your own free website on Tripod.com
Go to ScienceDirect Home Skip Main Navigation Links
 Register or Login:   Password:      
HomeSearchBrowse JournalsBrowse Abstract DatabasesBrowse Reference WorksMy AlertsMy ProfileHelp (Opens new window)
 Quick Search:  within Quick Search searches abstracts, titles, and keywords. Click for more information. 
Biological Conservation
Volume 106, Issue 3, August 2002, Pages 293-302
This Document
SummaryPlus
Full Text + Links
PDF (172 K)

Actions
Cited By
Save as Citation Alert
E-mail Article
Export Citation

DOI: 10.1016/S0006-3207(01)00255-5
PII: S0006-3207(01)00255-5
 
Copyright © 2002 Elsevier Science Ltd. All rights reserved.

Contrasting responses of plant and insect diversity to variation in grazing intensity

Andreas KruessCorresponding Author Contact Information, E-mail The Corresponding Author and Teja Tscharntke

Agroecology, Georg-August University, Waldweg 26, D-37075 Göttingen, Germany

Received 21 February 2001;  revised 27 October 2001;  accepted 1 November 2001.  Available online 7 December 2001.

Abstract

The effects of grazing intensity on plant and insect diversity were examined in four different types of grassland (intensively and extensively cattle-grazed pastures, short-term and long-term ungrazed grassland; 24 study sites). Vegetation complexity (plant species richness, vegetation height, vegetation heterogeneity) was significantly higher on ungrazed grasslands compared to pastures but did not differ between intensively and extensively grazed pastures. However, insect species richness was higher on extensively than on intensively grazed pastures, established by suction sampling of four insect taxa (Auchenorrhyncha, Heteroptera, Coleoptera, Hymenoptera Parasitica). This may be due to intensive grazing disrupting plant–insect associations as predicted by a "trophic-level" hypothesis. Local persistence and small-scale recolonization of insects on plants appeared to be difficult in the highly disturbed environment of intensive grazing. Insect diversity increased across the four treatments in the following order: intensively grazed<extensively grazed<short-term ungrazed<long-term ungrazed. The major predictor variable of differences in species diversity was found to be vegetation height. Predator–prey ratios within the investigated insect groups were not affected by grazing intensity.

Author Keywords: Grassland; Cattle grazing; Trophic-level interactions; Insect communities; Habitat management

Article Outline

1. Introduction
2. Methods
2.1. Study area
2.2. Characterization of the vegetation
2.3. Suction samples
2.4. Data analysis
3. Results
3.1. Vegetation
3.2. Insect samples
3.2.1. Auchenorrhyncha
3.2.2. Heteroptera
3.2.3. Coleoptera
3.2.4. Hymenoptera Parasitica
3.3. Insect species richness and abundance
4. Discussion
4.1. Vegetation
4.2. Insects
4.2.1. Intensively vs. extensively grazed pastures
4.2.2. Pastures vs. ungrazed grassland
4.3. Conclusions for conservation management
Acknowledgements
References

1. Introduction

In grassland habitats, insect diversity can be affected by habitat management since practices like cutting and grazing may change the associated insect community through profound alterations of plant growth, plant architecture and vegetation diversity (Strong and Huntly). Intense grazing can reduce arthropod species diversity and abundance ( Morris; Gibson; Rushton and McFerran) and relaxation of grazing pressure is known to increase species richness and abundance of phytophagous insects and their parasitoids ( Andrzejewski; Morris; Watts and Morris).

The effects of grazing on invertebrate diversity can be divided into short-term and long-term effects. Short-term effects are linked to (1) the simplification of plant architecture, for example the destruction of specific feeding niches, like seeds or tall-grass shoots (Andrzejewski; Morris; Morris; Hutchinson and Purvis), thereby reducing the insect diversity of taxonomic groups depending on such structures ( Andrzejewski; Lawton; Morris and Tscharntke), and (2) the rejuvenation of plant tissue due to regrowth of grazed plants, so that young and nutrient-rich plant-tissue is available ( Moore and Tscharntke). Long-term effects are caused by changes in the composition of the plant communities and thus in vegetation structure ( Day and Huntly).

Traditional types of management like low-intensity grazing or mowing once per annum are often associated with high biological diversity in anthropogenic habitats like grasslands (Tscharntke and Wettstein). The maintenance of these semi-natural habitats is the aim of various programmes created or supported by nature conservation agencies.

The aim of this study was to examine the success of reduced grazing intensity as a management tool to increase plant and insect biodiversity. Four different types of grassland representing a gradient of declining grazing intensity were compared: intensively grazed pastures>extensively grazed pastures>short-term ungrazed grassland>long-term ungrazed grassland.

We examined whether grazing effects can be attributed to one of the following hypotheses:

The diversity of phytophagous insects is positively correlated with plant species diversity, plant architectural diversity, occurrence of different vegetation patches and plant productivity (Strong; Siemann and Waide) and may therefore be enhanced by reduced grazing intensity.
The effects of consumers on their resources in both plant–herbivore and herbivore–enemy communities can be expected to increase with decreasing disturbance such as grazing intensity (Southwood; Tscharntke and Holt). This follows from the "trophic-level" hypothesis that trophic interactions such as plant–phytophagous insect interactions are interrupted by intense grazing because of separation of herbivores from their resource.

2. Methods

2.1. Study area

The investigated sites were situated in the "Bilsbek-Niederung" in Schleswig-Holstein, North Germany, 25 km north of Hamburg. The area is characterized by a heterogeneous landscape structure, dominated by grasslands and forests. The investigation, carried out between March and August 1996, covered four types of grassland habitat: (1) intensively grazed pastures, (2) extensively grazed pastures, (3) small fenced grassland areas centred on the extensive pastures, ungrazed for 3 years (short-term ungrazed), and (4) grassland, ungrazed for more than 5 years (long-term ungrazed). The four habitat types were each replicated six times (=24 study sites). The extensively grazed pastures were in management agreements with the Schleswig-Holstein Department of Nature Conservation as part of an extensification programme for pastures to support biodiversity. These agreements limited grazing intensity to 1.5 cattle per ha and permitted grazing between 1 May and 15 November.

Mean grazing intensity was 5.5±1.4 cattle/ha on intensively grazed pastures and 1.4±0.1 cattle/ha on extensively grazed pastures and differed significantly (ANOVA: F1,10=8.1, P=0.02, n=12).

To minimize the effects from other than the desired factors, the selected sites had to meet the following conditions: similarity in (1) former management, (2) area, and (3) location within the investigation area. Similarity in previous management was easy to meet since in the "Bilsbek-Niederung", there is a long tradition in cattle grazing on the grassland areas. The selected pastures had continuity in management intensity for at least four years. The short-term ungrazed areas were all established in spring 1993 on the extensively grazed pastures as an additional part of the extensification programme. It was not possible to evaluate the precise period without grazing for the long-term ungrazed grasslands due to changes of leaseholders, but we selected only sites that were known to be ungrazed for at least 5 years. To avoid bias due to local effects, we selected only those pastures included in the extensification programme with neighbouring intensively grazed pastures and long-term ungrazed grasslands. Thus, the final design consisted of six randomized blocks, each comprising one site of each of the four habitat types. Soil conditions and moisture were similar on all sites. Distances between the sites within each block were <1000 m, and distances between the blocks were 2–3 km. Mean area did not differ significantly between pastures (intensive: 2.9±0.4 ha, extensive: 3.6±1.1 ha) and long-term ungrazed grasslands (1.6±0.4 ha), but the short-term ungrazed grasslands were significantly smaller (0.02±0.002 ha).

2.2. Characterization of the vegetation

Vegetation measurements were recorded at three periods (within the first week of May, June, and July) in 1996. Vascular plant species were recorded during the whole season to obtain the total number of plant species per site. In May, the relative proportions of distinctive vegetation units >50 m2 were mapped for each site. For all plant species within these units, % cover was recorded. The vegetation units themselves differed in plant species composition or in abundance of dominant plant species (e.g. grass patches dominated by Holcus lanatus, patches dominated by Trifolium repens, thistle patches). The total number of vegetation units per site was counted as a measure of vegetation heterogeneity. For all patches, plant species composition and percent cover of each plant species were recorded (inside an area of 25 m2 in size). The area of each subsample (see later) was also characterized by vegetation height and percent cover of the vegetation. Nomenclature of plant species followed Ellenberg (1996).

2.3. Suction samples

Insects (Auchenorrhyncha, Heteroptera, Hymenoptera Parasitica, Coleoptera) were collected by suction sampling. Samples were taken monthly (four times from May to August) and, on each occasion, all treatments were sampled within 1 week. Each monthly sample consisted of eight subsamples taken in the four most abundant vegetation units (=highest percentage of the area) on each site. Thus, the insects were related to vegetation representative of 80–100% of the sites' total vegetation (intensively grazed: 90–100%; extensively grazed: 80–100%; short-term ungrazed: 100%; long-term ungrazed: 85–100%). On sites with only two (one intensively grazed site) or three (one extensively grazed site and two short-term ungrazed sites) vegetation units, sample sizes were completed by additional samples on these units. Each subsample was taken by covering a 0.25 m2 area of vegetation with a gauze-cage to prevent insects escaping. The insects inside the cage were then extracted into gauze bags using a UNIVAC portable suction sampler (JLO Type 35, Burkard Mfg. Co. Limited; vacuum 180 mm Hg, suction hose inner diameter 50 mm) for 5 min. The relatively long sampling period of 5 min per 0.25 m2 subsample was necessary to meet the problem of lower suction efficiency of the UNIVAC in dense and tall vegetation. Greiler (1994) achieved quantitative samples with the same method and only 3 min of sampling on set-aside areas and old meadows. The total area sampled per site (eight subsamples, 4 months) was 8 m2. Subsamples were frozen at -25 °C until insects were sorted and identified. This method allows quantitative comparisons between habitats (Arnold and Greiler). Dennis et al. (1998) described possible sources of error associated with the suction sampling method due to edge effects. Edge effects were minimized in this study since, as well as preventing insects from escaping from within the sampling area, the caging prevented specimens being drawn in from outside.

Auchenorrhyncha, Heteroptera and Coleoptera were identified to species by specialist taxonomists (see Acknowledgements), Hymenoptera Parasitica to "morphospecies". Information on the biology of species, especially host plant and habitat preferences, was collated from literature sources (Wagner; Wagner; Wagner; Remane; Freude; Ossiannilsson; Ossiannilsson; Ossianilsson; Gauld; Lohse; Lohse; Lohse and Goulet) and the taxonomists' comments.

2.4. Data analysis

Insect data and associated vegetation data were pooled for each site for further analyses, since the number of insect specimens collected in individual subsamples was often too low to allow a detailed sample analysis (e.g. by time and vegetation patch).

Data sets were log-transformed before analysis if they did not show a normal distribution. One-way ANOVA was carried out to test for treatment effects on vegetation and insect diversity. Stepwise multiple regression analyses were performed to identify the most important habitat and vegetation variables driving insect diversity in the investigated sites. Both ANOVA and stepwise regression analyses were performed with the data of three (intensively and extensively grazed pastures, long-term ungrazed grassland) out of the four grassland habitat types because data from short-term ungrazed sites and extensively grazed pastures were not independent (see Section 2.1). Therefore, an additional t-test for paired samples was performed to test for differences between extensively grazed pastures and short-term ungrazed areas. The variables offered to the multiple regression analyses are shown in Table 1. Arithmetic means are given ±1 standard error.

Table 1. List of habitat variables used in multiple regression analyses with species richness and abundance of the investigated insects set as dependent variables
View Table
(18K)

3. Results

3.1. Vegetation

Mean plant species richness per 25 m2 differed only slightly between intensively and extensively grazed pastures, short-term and long-term ungrazed grasslands, and differences were not significant (Fig. 1a). The total number of plant species found in the six replicates of the treatments was 121 species in intensively grazed pastures, 122 in extensively grazed pastures, 115 in the short-term ungrazed grasslands, and 176 species in the long-term ungrazed grasslands. In a comparison of the two types of pasture, 94 plant species were found in both treatments, 27 plant species were found only in the intensive treatment, and 28 plant species were exclusive to the extensive treatment.


Enlarge Image
(36K)

Fig. 1. (a)–(b) Vegetation diversity, (c)–(f) insect diversity from suction samples, in differently grazed pastures and ungrazed grasslands. I=intensively grazed pastures, E=extensively grazed pastures, S=short-term ungrazed grasslands, L=long-term ungrazed grasslands. (a) Mean plant species richness per 25 m2; (b) mean vegetation height per site; (c) total numbers of species and (d) total numbers of individuals of Coleoptera, Heteroptera, Auchenorrhyncha, morphospecies of parasitic Hymenoptera; (e) numbers of species and (f) numbers of individuals of grassland specialists (Coleoptera, Heteroptera, Auchenorrhyncha). Treatments with different letters above standard error bars are significantly different (P<0.01 from ANOVA for comparisons of I, E and L, n=18). * Indicates significant differences between E and S (P<0.05 from paired t-tests, n=12). All insect data were ln-transformed for analyses.

Mean vegetation height was significantly lower on grazed pastures compared to ungrazed grasslands, but did not differ significantly between intensively and extensively grazed pastures (Fig. 1b). Total vegetation cover ranged between 70% and 100% across the 24 sites and did not differ significantly between the four treatments.

The grass–herb species ratio was 0.46±0.02 on intensive, 0.50±0.03 on extensive pastures, 0.40±0.02 on the short-term and 0.45±0.04 on the long-term ungrazed grasslands. The grass–herb abundance ratio ranged from 5.9±2.6 on intensive and 4.2±1.2 on extensive pastures to 1.3±0.6 on short-term and 1.7±0.5 on long-term ungrazed grasslands. But differences between the four treatments were not significant for both ratios.

Vegetation heterogeneity (the number of different vegetation units) did not differ significantly (F2,15=1.9, P=0.18, n=18) between pastures (intensive: 4.8±0.7, extensive: 5.0±0.7) and long-term ungrazed grasslands (6.7±0.8). There was no significant correlation between vegetation heterogeneity and the habitat area of these three treatments (F1,16=0.21, r2=0.13, P=0.65, n=18), but in the very small short-term ungrazed grasslands, vegetation heterogeneity was significantly lower (3.7±0.2) compared with long-term ungrazed sites (F1,10=11.9, P=0.006, n=12).

At all 24 sites, vegetation was dominated by certain grasses (Elymus repens, Lolium perenne, Holcus lanatus, Phalaris arundinacea, Poa pratensis) and herbs (Cirsium arvense, Ranunculus repens, Trifolium repens, Urtica dioica). The abundance of individual plant species was significantly affected by grazing intensity in only a few cases. Two grass species dominated the pastures, Holcus lanatus and Lolium perenne. Abundance of the latter was significantly higher on intensively grazed pastures (25% cover) than on extensively grazed pastures (12% cover; F2,15=5.3, P=0.02, n=18) and the species was almost absent in the ungrazed grasslands. Two other abundant grasses, Agrostis capillaris (common bent grass) and Poa pratensis, showed contrasting responses to reduced grazing in the pastures. Agrostis capillaris suffered from intensive grazing, covering <1% in the intensive pastures but 12% in the extensive pastures (F2,15=4.8, P=0.02, n=18), whereas Poa pratensis benefited from intensive grazing, covering 11% in the intensive, but <1% in the extensive pastures (F2,15=5.6, P=0.02, n=18).

Two herbaceous species also showed significant changes in abundance. Urtica dioica covered 3% in the pastures but had a higher abundance in the long-term ungrazed grasslands (47%; F2,15=14.3, P<0.001, n=18) and also in the short-term ungrazed sites (34%). In contrast, the abundance of Trifolium repens was significantly higher in intensively (4% cover) and extensively (6% cover) grazed pastures than in grasslands (1% cover; F2,15=4.5, P=0.03, n=18).

3.2. Insect samples

From May to August a total of 5663 individuals were collected by suction sampling. The most abundant orders were Auchenorrhyncha and parasitic Hymenoptera. Species richness was highest in Coleoptera and parasitic Hymenoptera and lowest in Heteroptera.

3.2.1. Auchenorrhyncha

A total of 3637 individuals, comprising 60 species, were sampled. Most of the species had a very low abundance (31 species with less-than-or-equals, slant10 individuals, 20 species with >10 and less-than-or-equals, slant 100 individuals, nine species with >100 individuals). All abundant species were grassland species, with a well-known preference for wet and nutritious habitats. The most abundant species were the Cicadellidae Arthaldeus pascuellus (Féll.), a wide-spread species of nutritious grassland, Streptanus sordidus (Zett.), Steptanus aemulans (Kirchb.), Errastunus ocellaris (Fáll.), and Macrostelis sp. All these species preferred ungrazed grasslands, whereas Macrostelis sp. was mainly associated with extensively grazed pastures.

3.2.2. Heteroptera

A total of 49 species and 274 individuals were sampled. Forty-three species were found as <10 individuals, while the others ranged in abundance from 10 to 37 individuals. The most numerous species was the predator Nabis limbatus (Dahlb.) (Nabidae). The most abundant phytophagous species were Leptoterna dolobrata (L.), Pithanus maerkeli (H.-Sch.) (Miridae), and Ichnodemus sabuleti (Fáll.) (Lygaeidae). Heteroptera species diversity and abundance was dominated by phytophagous species. The percentage of herbivores was 62.2% on long-term and 64.0% on short-term ungrazed grasslands, 68.9% on intensively grazed and 83.8% on extensively grazed pastures, but these differences were not statistically significant.

3.2.3. Coleoptera

A total of 115 species and 608 individuals were recorded. Of these, 104 species had a very low abundance of <10 individuals. Three of the most numerous species were the nettle specialists Brachypterus urticae (F.) and Brachypterus glaber Steph. (Kateredidae), and the clover weevil Ischnopterapion virens (Hbst.) (Apionidae). Two predacious ladybirds, Coccinella 7-punctata L. and Propylea 14-guttata (L.) (Coccinellidae) were also abundant. The percentage of phytophagous specimens was 41.4% in extensively and 45.8% in intensively grazed pastures, 57.2% in short-term and 60.8% in long-term ungrazed grasslands, but these differences were not statistically significant.

3.2.4. Hymenoptera Parasitica

A total of 1144 specimens, comprising 105 morphospecies, were sampled. Chalcidoidea were the most abundant group, comprising 519 specimens of 53 morphospecies. The second and third most abundant groups were Ichneumonoidea (505 specimens comprising 38 morphospecies), and Proctotrupoidea (91 specimens, 10 morphospecies).

3.3. Insect species richness and abundance

As shown in Fig. 1c, species richness was significantly lower on intensively grazed pastures than on extensively grazed pastures and on ungrazed grasslands. Insect abundance was lowest in the intensively grazed pastures and significantly higher on ungrazed grasslands ( Fig. 1d). In contrast to species richness, abundance on the extensively grazed pastures did not significantly differ from abundance on intensively grazed pastures.

A similar effect of grazing intensity on species richness and abundance was found in the distribution of insects known to be grassland specialists. Pooled species richness and abundance of grassland specialists from Auchenorrhyncha, Heteroptera and Coleoptera was more than eight-fold higher on long-term ungrazed grasslands than on intensively grazed pastures (Fig. 1e, f). As for the total insects above, species richness (but not the abundance) of grassland specialists was significantly higher on extensively than on intensively grazed pastures. On the short-term ungrazed grasslands abundance, but not species richness, of grassland specialists was significantly lower than on the surrounding extensively grazed pastures.

These patterns of insect distribution among the treatments were supported by further ANOVA and t-test analyses of feeding guilds and habitat specialists (Table 2). The results allow the following conclusions. Firstly, species richness was significantly higher in extensive than intensive pastures in only three (predacious species, total species of Coleoptera, parasitic Hymenoptera) out of 18 cases. Abundance of insects was significantly higher in four cases (predacious specimens, Coleoptera, predacious Coleoptera, Hymenoptera Parasitica). Secondly, insect diversity was highest on the long-term ungrazed grasslands in almost all cases (with the exception of species richness of Heteroptera feeding guilds). Thirdly, insect species richness and abundance on the short-term ungrazed grasslands were at least on the same level as in the surrounding extensively grazed pastures, and in one case (predatory Heteroptera) it was even higher (Table 2). Fourthly, predator–prey ratios did not differ significantly among the four treatments when calculated for both number of species and number of specimens.

Table 2. Differences in insect species richness and abundance between the four treatmentsa
View Table
(46K)

Stepwise multiple regression analyses were performed to analyze whether insect species richness and abundance are affected by habitat or vegetation characteristics listed in Table 1. The results of these analyses are summarized in Table 3. Vegetation height was the first or only variable explaining both species richness and abundance in about 80% of the multiple regression analyses performed across all investigated taxa. In all these cases, species richness and abundance were positively correlated with vegetation height, indicating that this is a very strong pattern (Table 3). Furthermore, increasing vegetation heterogeneity enhanced both the total abundance of insects and the abundance of phytophagous insects, particularly the Auchenorrhyncha. Greater plant species richness enhanced the species richness of grassland specialists and mono- or oligophagous herbivores, particularly the Coleoptera. Heteroptera were the least numerous group, both in terms of individuals and species. Heteroptera abundance was positively correlated with vegetation height. Both the number of morphospecies and general abundance of parasitic Hymenoptera were also positively correlated with vegetation height, presumably due to the strong positive relationship between vegetation height and the diversity of potential hosts.

Table 3. Results from multiple regression analyses of insect species richness and abundance (treatments I, E and L; n=18) with habitat variablesa
View Table
(37K)

4. Discussion

4.1. Vegetation

The results of this study showed that reduction in grazing intensity enhanced insect but not plant diversity. Although grazing intensity in extensively grazed pastures was 75% lower than in intensively grazed pastures, neither plant species diversity nor vegetation heterogeneity or vegetation height differed significantly. The expectation that grazing at low stocking rates may create a mosaic of heavily and lightly grazed areas (WallisDeVries and Raemakers, 2001) was also not supported by the measure of vegetation heterogeneity. Only a few grasses (Agrostis capillaris, Lolium perenne, Poa pratensis) and herbs (Trifolium repens, Urtica dioica) that are well-known to respond to grazing reduction showed significant differences in abundance between intensively and extensively grazed pastures. Plant species like Lolium perenne, Poa pratensis, and Trifolium repens, which are known to be highly tolerant to grazing and trampling (Cole, 1995), or less competitive (Poa pratensis, Wilson and Tilman, 1991) were favoured by intensive grazing. In contrast, grasses with a low regrowth capability such as Agrostis capillaris (Crawley, 1983) gained from grazing reduction.

4.2. Insects

Differences in insect species richness and abundance between intensively and extensively grazed pastures and ungrazed grasslands showed a very consistent pattern for all insect taxa. As a rule, insect diversity was lowest on intensively grazed pastures and was highest on abandoned land (Fig. 1c, d).

4.2.1. Intensively vs. extensively grazed pastures

Insect diversity was about 50% lower on intensively compared with extensively grazed pastures, whereas vegetation diversity (plant species richness, vegetation heterogeneity, vegetation height, grass–herb ratios) was not significantly affected. The significant differences in insect species richness between intensively and extensively grazed pastures cannot be attributed to a "resource heterogeneity" or "resource productivity" hypothesis, because plant species richness, vegetation height, number of differential vegetation units, and vegetation cover were not different between intensively and extensively grazed pastures. Furthermore, insect abundance was similar among the pastures, so the higher number of species was not due to more sampled individuals. Thus, our results on intensively and extensively grazed pastures differ from former studies, where differences in insect diversity were simply attributed to differences in plant taxonomic diversity, plant productivity (Walsingham; Southwood; Lawton; Brown; Rosenweig and Siemann), plant architecture ( Andrzejewski; Brown and Lawton) and the destruction of feeding niches ( Andrzejewski, 1965). Our results support only the "trophic-level" hypothesis as the fourfold cattle density appeared to affect the plant–insect associations by disrupting the continuity of feeding by phytophagous insects. Presumably, insects of low abundance should experience most difficulty in persisting locally in a highly disturbed environment. Moreover, small-scale recolonization may be limited by the dispersal capabilities of different insects ( Cornell and Hanski).

Our results from suction samples suggest that grazing intensity of pastures affected phytophagous and entomophagous insects equally. But evidence is limited because important predator groups (i.e. spiders) were not considered.

4.2.2. Pastures vs. ungrazed grassland

The effects of grazing reduction on insect diversity were most obvious when pastures and grasslands were compared. There was a general trend of higher diversity on ungrazed sites compared to the pastures. Whereas this was only a slight tendency for short-term abandonment, these differences were significant for most insect taxa, feeding guilds, and habitat specialists on long-term abandoned grasslands. Ungrazed grasslands supported a higher resource heterogeneity than pastures in that vegetation height was greater and, thereby, plant architecture more complex (Murdoch and Southwood). However, plant species richness and vegetation heterogeneity, which are usually expected to be of major importance, did not show any differences. The "resource productivity" hypothesis was supported in that vegetation height differed (on the basis of similar measures of vegetation cover). Last but not least, ungrazed grasslands are not affected by cattle disturbance, so plant–insect associations are not disrupted.

The results of the multiple regression analyses across all 18 independent study sites emphasized the general importance of vegetation height for the prediction of species richness and abundance of insects. The finding that vegetation height is the most important variable for explaining insect diversity and abundance is in support of a study with Lepidoptera on Calluna vulgaris (Haysom and Coulson, 1998). In contrast to other studies ( Morris and Morris) we did not find that Auchenorrhyncha responded more strongly than Coleoptera to simplification of the vegetation structure by grazing. In contrast, the observed differences in species richness and abundance between intensively and extensively grazed pastures were >50% in Coleoptera, but only 20–30% in Auchenorrhyncha. The number of rare species usually increases with total insect abundance ( Rosenzweig, 1995). Rare (<10 specimens) species dominated the insect community in this and other studies (see Greiler and Tscharntke). Insect species richness may therefore be supported indirectly by increasing plant productivity, since arthropod abundance is positively dependent on plant productivity, and increasing total abundance may encourage rarer species to persist locally ( Hutchinson; Brown and Rosenweig).

In conclusion, we observed no change in plant diversity, but an increase in insect diversity following a reduction in grazing intensity. In contrast to expectations, grazing or abandonment did not affect habitat specialists (grassland species) more than generalists. A "bottom-up" effect of enhanced herbivore diversity due to a corresponding increase of vegetation diversity could not be found, supporting a "trophic-level" hypothesis that intense cattle grazing may disrupt plant–insect interactions. However, predators and parasitoids were no more supported than herbivores, so predator–prey ratios were very similar among the different treatments. Thus, reduced grazing intensity and grassland abandonment enhanced the second, but did not appear to enhance the relative importance of the third trophic level.

4.3. Conclusions for conservation management

With respect to habitat management and biological conservation we conclude a reduction in grazing intensity on grassland may successfully enhance insect diversity, even when plant diversity does not show any changes. The strongest effects on insect diversity were seen on the large and long-term ungrazed habitats, but there were also positive effects of small-sized and short-term ungrazed areas. This underlines the importance of small-scale undisturbed areas for the preservation and restoration of insect diversity. Since the long-term protection of grassland and early successional fallows depends on management such as grazing, cutting or burning, only extensive use of grassland and fallows can maintain these habitats and their biodiversity.

Acknowledgements

Helpful comments by Brian Davis, Karen Haysom, Peter Dennis, Carsten Thies and two anonymous referees greatly improved this manuscript. Thanks for identification of the insects go to Boris Büche (Coleoptera), Albert Melber (Heteroptera) and Herbert Nickel (Auchenorrhyncha). Thanks to Carsten Scheel and Holger Schmidt for carrying out suction samples and insect sorting. Thanks also to Gisela Bertram for vegetation mapping. Financial support came from the Landesamt für Natur und Umwelt Schleswig-Holstein.

References

Andrzejewski, 1965. L. Andrzejewski , Stratification and its dynamics in meadow communities of Auchenorrhyncha (Homoptera). Ekologika Polska 13 (1965), pp. 685–715.

Andrzejewski and Gyllenberg, 1980. L. Andrzejewski and G. Gyllenberg , Small herbivore subsystem. In: A.I. Breymeyer and G.M. van Dyne, Editors, Grasslands, System Analysis and Man, Cambridge University Press, Cambridge (1980), pp. 201–267.

Arnold et al., 1973. A.J. Arnold, P.H. Needham and J.H. Stevenson , A self-powered portable suction sampler and its use to assess the effects of azinos methyl and endosulfan on blossom beetle populations on oil seed rape. Annals of Applied Biology 75 (1973), pp. 229–233.

Brown, 1981. J.H. Brown , Two decades of homage to Santa Rosalia: towards a general theory of diversity. American Zoologist 21 (1981), pp. 877–888.

Brown, 1991. V.K. Brown , The effect of changes in habitat structure during succession in terrestrial communities. In: S.S. Bell, E.D. McCoy and H.R. Muhinsky, Editors, Habitat Structure: the Physical Arrangement of Objects in Space, Chapman and Hall, London (1991), pp. 141–168. Abstract-GEOBASE  

Brown et al., 1992. V.K. Brown, C.W.D. Gibson and J. Kathirithamby , Community organisation in leaf hoppers. Oikos 65 (1992), pp. 97–106. Abstract-GEOBASE  

Cole, 1995. D.N. Cole , Experimental trampling of vegetation. 2. Predictors of resistance and resilience. Journal of Applied Ecology 32 (1995), pp. 215–224. Abstract-Elsevier BIOBASE | Abstract-GEOBASE  

Cornell and Lawton, 1992. H.V. Cornell and J.H. Lawton , Species interactions, local and regional processes, and limits to the richness of ecological communities––a theoretical perspective. Journal of Animal Ecology 61 (1992), pp. 1–12. Abstract-GEOBASE  

Crawley, 1983. M.J. Crawley , Herbivory: the Dynamics of Animal–Plant Interactions. , Blackwell Scientific Publications, Oxford (1983).

Day and Detling, 1990. T.A. Day and J.K. Detling , Grassland patch dynamics and herbivore grazing preference following urine deposition. Ecology 71 (1990), pp. 180–188. Abstract-GEOBASE  

Dennis et al., 1998. P. Dennis, M.R. Young and I.J. Gordon , Distribution and abundance of small insects and arachnids in relation to structural heterogeneity of grazed, indigenous grasslands. Ecological Entomology 23 (1998), pp. 253–264. Abstract-Elsevier BIOBASE | Abstract-GEOBASE   | Full Text via CrossRef

Ellenberg, 1996. H. Ellenberg , Vegetation Mitteleuropas mit den Alpen in ökologischer, dynamischer und historischer Sicht. , Eugen Ulmer, Stuttgart (1996).

Freude et al., 1964–1983. Freude, H., Harde, K.W., Lohse, G.A., 1964–1983. Die Käfer Mitteleuropas, 1–11. Goecke & Evers, Krefeld.

Gauld and Bolton, 1988. I. Gauld and B. Bolton , The Hymenoptera. , Oxford University Press, Oxford (1988).

Gibson et al., 1992. C.W.D. Gibson, V.K. Brown, L. Losito and G.C. McGavin , The response of invertebrate assemblies to grazing. Ecography 15 (1992), pp. 166–176. Abstract-GEOBASE  

Goulet and Huber, 1993. H. Goulet and J.T. Huber , Hymenoptera of the World: An Identification Guide to Families. , Agriculture Canada Publication, Ottawa (1993).

Greiler, 1994. H.-J. Greiler , Insektengesellschaften auf selbstbegrünten und eingesäten Ackerbrachen. Agrarökologie 11 (1994), pp. 1–136.

Hanski and Gyllenberg, 1997. I. Hanski and M. Gyllenberg , Uniting two general patterns in the distribution of species. Science 275 (1997), pp. 397–400. Abstract-GEOBASE | Abstract-GEOBASE   | Full Text via CrossRef

Haysom and Coulson, 1998. K.A. Haysom and J.C. Coulson , The Lepidoptera fauna associated with Calluna vulgaris: effects of plant architecture on abundance and diversity. Ecological Entomology 23 (1998), pp. 377–385. Full Text via CrossRef

Holt et al., 1999. R.D. Holt, J.H. Lawton, G.A. Polis and N.D. Martinez , Trophic rank and species–area relationship. Ecology 80 (1999), pp. 1495–1504. Abstract-Elsevier BIOBASE | Abstract-GEOBASE  

Huntly, 1991. N. Huntly , Herbivores and the dynamics of communities and ecosystems. Annual Reviews of Ecology and Systematics 22 (1991), pp. 477–504.

Hutchinson, 1959. G.E. Hutchinson , Homage to Santa Rosalia or why are there so many kinds of animals. American Naturalist 93 (1959), pp. 145–159. Full Text via CrossRef

Hutchinson and King, 1980. K.J. Hutchinson and K.L. King , The effects of sheep stocking level on invertebrate abundance, biomass and energy utilization in a temperate, sown grassland. Journal of Applied Ecology 17 (1980), pp. 369–387.

Lawton, 1983. J.H. Lawton , Plant architecture and the diversity of phytophagous insects. Annual Review of Entomology 28 (1983), pp. 23–39. Abstract-GEOBASE  

Lawton and Schoder, 1977. J.H. Lawton and D. Schröder , Effects of plant type, size of geographical range, and taxonomic isolation on number of insect species associated with British plants. Nature 265 (1977), pp. 137–140.

Lohse and Lucht, 1989. G.A. Lohse and W.H. Lucht , Die Käfer Mitteleuropas, Supplementband 1. , Goecke & Evers, Krefeld (1989).

Lohse and Lucht, 1992. G.A. Lohse and W.H. Lucht , Die Käfer Mitteleuropas, Supplementband 2. , Goecke & Evers, Krefeld (1992).

Lohse and Lucht, 1994. G.A. Lohse and W.H. Lucht , Die Käfer Mitteleuropas, Supplementband 3. , Goecke & Evers, Krefeld (1994).

McFerran et al., 1994. D.M. McFerran, W.I. Montgomery and J.H. McAdam , The impact of grazing on communities of ground-dwelling spiders (Araneae) in upland vegetation types. Biology and Environment: Proceedings of the Royal Irish Academy 94B 2 (1994), pp. 119–126.

Moore and Clements, 1984. D. Moore and R.O. Clements , Stem-borer larval infestation of ryegrass swards under rotationally grazed and cut conditions. Journal of Applied Ecology 21 (1984), pp. 581–590. Abstract-GEOBASE  

Morris, 1967. M.G. Morris , Differences between the invertebrate faunas of grazed and ungrazed chalk grasslands. I. Responses of some phytophage insects to a cessation of grazing. Journal of Animal Ecology 36 (1967), pp. 459–474.

Morris, 1981. G.M. Morris , Responses of grassland invertebrates to management by cutting. 3. Adverse effects on Auchenorrhyncha. Journal of Applied Ecology 18 (1981), pp. 107–123.

Morris and Lakhani, 1979. M.G. Morris and K.H. Lakhani , Responses of grassland invertebrates to management by cutting. Journal of Applied Ecology 16 (1979), pp. 77–98. Abstract-MEDLINE  

Morris and Plant, 1983. M.G. Morris and R. Plant , Responses of grassland invertebrates to management by cutting. 5.Changes in Hemiptera following cessation of management. Journal of Applied Ecology 20 (1983), pp. 157–177. Abstract-GEOBASE  

Morris and Rispin, 1987. G.M. Morris and W.E. Rispin , Abundance and diversity of the coleopterous fauna of a calcarous grassland under different cutting regimes. Journal of Applied Ecology 24 (1987), pp. 451–466.

Murdoch et al., 1972. W.W. Murdoch, F.C. Evans and C.H. Peterson , Diversity and pattern in plants and insects. Ecology 53 (1972), pp. 819–829.

Ossiannilsson, 1978. Ossiannilsson, F., 1978. The Auchenorrhyncha (Homoptera) of Fennoscandia and Denmark. Part 1: Introduction, infraorder Fulgomorpha. Fauna Entomologica Scandinavica 7 (1). Scandinavia Science Press, Klampenborg.

Ossiannilsson, 1981. Ossiannilsson, F., 1981. The Auchenorrhyncha (Homoptera) of Fennoscandia and Denmark. Part 2: The families Cicadidae, Cercopidae, Membracidae, Cicadellidae (excl. Deltocephalinae). Fauna Entomologica Scandinavica 7 (2). Scandinavia Science Press, Klampenborg.

Ossianilsson, 1983. Ossiannilsson, F., 1983. The Auchenorrhyncha (Homoptera) of Fennoscandia and Denmark. Part 3: The family Cicadellidae. Fauna Entomologica Scandinavica 7 (3). Scandinavia Science Press, Klampenborg.

Purvis and Curry, 1981. G. Purvis and J.P. Curry , The influence of sward management on foliage arthropod communities in a ley grassland. Journal of Applied Ecology 18 (1981), pp. 711–725.

Remane, 1958. R. Remane , Die Besiedlung von Grünlandflächen verschiedener Herkunft durch Wanzen und Zikaden im Weser-Ems-Gebiet. Zeitschrift für Angewandte Entomologie 42 (1958), pp. 353–400.

Rosenweig, 1995. M.L. Rosenzweig , Species Diversity in Space and Time. , Cambridge University Press, Cambridge (1995).

Rosenweig and Abramsky, 1993. M.L. Rosenzweig and Z. Abramsky , How are productivity and diversity related? Species diversity in ecological communities. In: E.R. Ricklefs and D. Schluter, Editors, Historical and Geographical Perspectives, Chicago University Press, Chicago (1993), pp. 52–65. Abstract-GEOBASE  

Rushton and Eyre, 1992. S.P. Rushton and M.D. Eyre , Grassland spider habitats in Northeast England. Journal of Biogeography 19 (1992), pp. 99–108. Abstract-GEOBASE  

Siemann, 1998. E. Siemann , Experimental test of effects of plant productivity and diversity on grassland arthropod diversity. Ecology 79 (1998), pp. 2057–2070. Abstract-GEOBASE  

Southwood, 1988. T.R.E. Southwood , Tactics, strategies and templets. Oikos 52 (1988), pp. 3–18. Abstract-GEOBASE  

Southwood et al., 1979. T.R.E. Southwood, V.K. Brown and P.M. Reader , The relationships of plant and insect diversities in succession. Biological Journal of the Linnean Society 12 (1979), pp. 327–348.

Strong et al., 1984. D.R. Strong, J.H. Lawton and T.R.E. Southwood , Insects on Plants. Community Patterns and Mechanisms. , Blackwell Scientific Publications, Oxford (1984).

Tscharntke, 1997. T. Tscharntke , Vertebrate effects on plant–invertebrate food-webs. In: A.C. Gange and V.K. Brown, Editors, Multitrophic Interactions in Terrestrial Ecosystems. Proc. 36th Symp. Brit. Ecol. Soc., Blackwell Science Ltd, Oxford (1997), pp. 277–297.

Tscharntke and Greiler, 1995. T. Tscharntke and H.-J. Greiler , Insect communities, grasses, and grasslands. Annual Review of Entomology 40 (1995), pp. 535–558. Abstract-GEOBASE | Abstract-Elsevier BIOBASE  

Wagner, 1952. Wagner, E., 1952. Blindwanzen oder Miriden. In: Dahl, F. (Ed.), Die Tierwelt Deutschlands, 54. Teil. Gustav Fischer Verlag, Jena.

Wagner, 1966. Wagner, E., 1966. Wanzen oder Heteropteren. I. Pentatomorpha. In: Dahl, F. (Ed.), Die Tierwelt Deutschlands, 54. Teil. Gustav Fischer Verlag, Jena.

Wagner, 1967. Wagner, E., 1967. Wanzen oder Heteropteren. II. Cimicimorpha. In: Dahl, F. (Ed.), Die Tierwelt Deutschlands, 55. Teil. Gustav Fischer Verlag, Jena.

Waide et al., 1999. R.B. Waide, M.R. Willig, C.F. Steiner, G. Mittelbach, L. Gough, S.I. Dodson, G.P. Juday and R. Parmenter , The relationship between productivity and species richness. Annual Reviews of Ecology and Systematics 30 (1999), pp. 257–300. Abstract-Elsevier BIOBASE | Abstract-GEOBASE   | Full Text via CrossRef

WallisDeVries and Raemakers, 2001. M.F. WallisDeVries and I. Raemakers , Does extensive grazing benefit butterflies in coastal dunes. Restoration Ecology 9 (2001), pp. 179–188. Abstract-GEOBASE   | Full Text via CrossRef

Walsingham, 1978. J.M. Walsingham , Effect of sheep grazing on the invertebrate population of agricultural grassland. Scientific Proceedings of the Royal Dublin Society Series A 6 (1978), pp. 297–304.

Watts et al., 1982. J.G. Watts, E.W. Huddleston and J.C. Owens , Rangeland Entomology. Annual Review of Entomology 27 (1982), pp. 283–311.

Wettstein and Schmid, 1999. W. Wettstein and B. Schmid , Conservation of arthropod diversity in montane wetlands: effect of altitude, habitat quality and habitat fragmentation on butterflies and grasshoppers. Journal of Applied Ecology 36 (1999), pp. 363–373. Abstract-Elsevier BIOBASE | Abstract-GEOBASE   | Full Text via CrossRef

Wilson and Tilman, 1991. S.D. Wilson and D. Tilman , Components of plant competition along an experimental gradient of nitrogen availability. Ecology 72 (1991), pp. 1050–1065. Abstract-GEOBASE  

Corresponding Author Contact Information Corresponding author. Tel.: +49-551-392-359; fax: +49-551-398-806; email: akruess@gwdg.de
This Document
SummaryPlus
Full Text + Links
PDF (172 K)

Actions
Cited By
Save as Citation Alert
E-mail Article
Export Citation
Biological Conservation
Volume 106, Issue 3, August 2002, Pages 293-302


 
HomeSearchBrowse JournalsBrowse Abstract DatabasesBrowse Reference WorksMy AlertsMy ProfileHelp (Opens new window)

Send feedback to ScienceDirect
Software and compilation © 2002 ScienceDirect. All rights reserved.
ScienceDirect® is an Elsevier Science B.V. registered trademark.


Your use of this service is governed by Terms and Conditions. Please review our Privacy Policy for details on how we protect information that you supply.