Several arthropods are known to be highly beneficial to agricultural production. Consequently it is of great relevance to study the importance of land management and land composition for the conservation of beneficial aphid-predator arthropod species in agricultural areas. Therefore our study focusing on the beneficial arthropod Bembidion lampros had two main purposes: I) identifying the physical barriers to the species' dispersal in the agricultural landscape, and II) assessing the effect of different land management strategies (i.e. use of pesticides and intensiveness) on the dispersal patterns. The study was conducted using genetic analysis (microsatellite markers) applied to samples from two agricultural areas (in Denmark) with different agricultural intensity. Land management effects on dispersal patterns were investigated with particular focus on: physical barriers, use of pesticide and intensity of cultivation.
The results showed that Bembidion lampros disperse preferably through hedges rather than fields, which act as physical barriers to gene flow. Moreover the results support the hypothesis that organic fields act as reservoirs for the re-colonization of conventional fields, but only when cultivation intensity is low. These results show the importance of non-cultivated areas and of low intensity organic managed areas within the agricultural landscape as corridors for dispersal (also for a species typically found within fields). Hence, the hypothesis that pesticide use cannot be used as the sole predictor of agriculture's effect on wild species is supported as land structure and agricultural intensity can be just as important.
Citation: Marchi C, Andersen LW, Loeschcke V (2013) Effects of Land Management Strategies on the Dispersal Pattern of a Beneficial Arthropod. PLoS ONE 8(6): e66208. https://doi.org/10.1371/journal.pone.0066208
Editor: Anna-Liisa Laine, University of Helsinki, Finland
Received: December 18, 2012; Accepted: May 7, 2013; Published: June 11, 2013
Copyright: © 2013 Marchi et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: Funded by the Ministry of Food, Agriculture and Fisheries under the Finance and Appropriation Act, Sections 24.33.02.10. Part of the Ph.D. project is funded by AGSoS (Aarhus Graduate School of Science). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
A number of arthropods, typically living in agricultural areas, have been shown to be highly beneficial to agricultural production, because they act as predators on pest species . These beneficial species are of major importance especially in relation to organic farming, and their abundance and dispersal pattern can be expected to be influenced by land management.
Bembidion lampros, a small polyphagus arthropod (of the Carabidae family) has been shown to be efficient as aphid pest controller in agricultural areas – and can also be used as indicator for insect abundance and diversity . Effects of landscape structure and land use on B. lampros have been widely investigated, but focusing only on the species abundance rather than on the drivers of dispersal in the agricultural landscape.
The typical B. lampros life cycle is marked by an overwintering stage where individuals reside in hedges at the field margins and a dispersal stage in early spring where they move into the field . The timing of dispersal is determined by air temperature and occurs in Denmark usually between early March and early April . After reproducing in the field the individuals return in early autumn to the hedges to overwinter , . B. lampros is univoltine  and short-lived but it lives mostly more than a year. It is one of the early spring-moving species , thus it has probably a relatively larger effect on aphid pest control compared to later-moving species . While its life cycle has been thoroughly studied, the dispersal routes within the agricultural land are largely unknown. In particular it is not known whether B. lampros cross hedgerows, tend to return to the same hedgerow after reproduction or easily cross ploughed fields. Such knowledge however is crucial for understanding the population dynamics of this highly beneficial species . Since this information is difficult to get through mark-recapture studies (given the very high population number and the small size of this species), the use of molecular genetic markers may help to elucidate this part of its ecology. Knowing which factors act as barrier to dispersal can be used to predict the probability of an area to receive immigrants and consequently be re-colonized in case of a sudden drop in abundance (e.g. after pesticides spraying). This knowledge could allow the implementation of management strategies that prevent populations going extinct or suffer severe bottlenecks (which could then lead to local extinction).
Hedgerow types, soil type, presence of pesticides and number of mechanical soil treatments have a clear effect on the abundance of B. lampros –. In particular many commonly used insecticides are shown to be detrimental to this species, both under laboratory and field conditions , . Even pesticides legally used in organic fields in many European countries (but not in Denmark) have been proven to be harmful . The use of pesticides in the field has led to a reduction in B. lampros abundance persisting up to four weeks after spraying; the following return to abundance levels preceding the spraying was ascribed to an effect of re-colonization from untreated nearby areas . This might suggest that fields grown organically can act as genetic reservoirs for the re-colonization of conventional fields.
Several studies have shown the importance of cultivation intensity for the abundance and mobility of B. lampros, in particular of low yield cultivation with high weed abundance , . Therefore the effect of agricultural intensiveness was included in our study by comparing two areas with very different agricultural intensity. The organic fields in the two areas also differed in time since conversion from conventional fields.
The main objectives of our study were :
- to analyze the population structure and gene flow pattern in the agricultural landscape and thereby identifying possibly dispersal barriers (with particular attention to fields, hedgerows and roads).
- to test the hypothesis that organically grown fields act as genetic reservoirs for conventional grown fields and the hypothesis that areas with low intensity agricultural land use have a higher genetic diversity, effective population size and comparatively higher gene flow.
Materials and Methods
We selected two agricultural areas with different intensity of cultivation composed of conventional and organic fields separated by tree-covered hedgerows. Sampling was performed for two consecutive years in early autumn (when individuals return to the hedgerows to overwinter) at sampling sites along the hedgerows.
Individuals were sampled along the hedgerows from two agricultural areas in Denmark, Kalø 56.363° North 9.700° East, and Bjerringbro 56.290° Nord 10.500° East. The two areas represented an intensive (Bjerringbro) and an extensive (Kalø) type of management: Bjerringbro area had a higher average amount of cultivated areas (66%) compared to the national level (61%), while the opposite was true for the Kalø area where only 52% of land is used for agriculture (Beate Strandberg, pers. comm.). In Kalø, the fields have been organically managed for many years (>15 years), longer than in Bjerringbro, and the weed cover in the fields in Kalø was also much higher (Beate Strandberg pers. comm.). The sampling scheme was designed to obtain samples from hedgerows dividing two organic or two conventional fields as well as from hedgerows dividing organic and conventional fields (on both sides). Regarding the pesticide regime the organic fields were never treated with pesticides/herbicides while these were routinely applied to conventional fields. Therefore conventional vs. organic classification defines pesticide regime. Each hedgerow presented four sampling sites 50 meters apart, two on both sites just opposite each other with a width of one meter. A total of 1140 samples were hand-picked during the two seasons (see Table 1 and Fig. 1 for details). Before DNA-extraction all individuals were identified to species morphologically.
The top map shows the location of the two areas used in this study. The maps at the bottom show the location and name of each sampling site for both areas (B = Bjerringbro, K = Kalø) and management types of the field (C for conventional or O for organic) and ordinal number.
DNA extraction and PCR details
Samples were frozen dry upon collection and stored at −20°C. DNA was extracted using a standard CTAB buffer and proteinase-K procedure , . Two multiplex PCR runs were used to amplify 15 microsatellite markers developed specifically for this species (Marchi et al. unpublished) using Qiagen Multiplex enzyme following the manufactories instructions. First multiplex: BL8, BL11, BL19, BL22, BL28, BL28b, BL30, BL34; second multiplex: BL2, BL5, BL6, BL12, BL14, BL24, BL27). Thermal profile, for both multiplex runs, was: 95°C for 15 min; then 35 cycles of 30 s at 94°C, 45 s at 57°C and 60 s at 72°C, with a final extension at 60°C for 30 min. PCR products were analysed using an ABI 3730 automated sequencer and typed using GeneMapper version 4.1 (Applied Biosystems).
Genetic diversity and effective population size analysis
Genetic diversity for each sampling site and each year was estimated as allelic richness with fstat  and expected heterozygosity with GenAlEx . Tests for goodness of fit to Hardy-Weinberg expectations and linkage equilibrium were performed in fstat. Moreover effective population size (Ne) was estimated with NeEstimator  for each sampling site for both years using “heterozygote excess method”  and Waples'  moments based approach. The differences in genetic diversity and Ne among conventional and organic sampling sites were tested using a Student t-test. The same tests were repeated, grouping the samples with regard to neighboring field's management as well (sampling site in conventional field with neighboring organic field versus sampling site in conventional field with neighboring conventional field and so on for all pairs).
Genetic structure analysis
Population structure was evaluated using a Bayesian based cluster analysis implemented in geneland R package . The analyses were performed for Kalø and Bjerringbro data using geographic coordinates as prior and k(max) = 10 (based on five separate runs each with 20,000/100,000 burn-in/sampling iterations).
The population structure was also estimated using pairwise multilocus Fst between sampling sites using arlequin v. 3.5 .
Gene flow analysis
The effect of environmental variables on the pattern of gene flow between sampling sites were assessed using bimr . bimr estimates recent gene flow to find the best explanatory factors for the recovered pattern among the given environmental distance matrices using a linear model fit with the gene flow matrix acting as dependent variable (a null model, including none of the specified factors, is always included in the analysis). The variables included in the analysis were: distance, hedgerow of belonging, field of belonging (hedgerow or field nearest to the sampling site), management type (conventional or organic) and road side (only for Bjerringbro). The inclusion of distance between sampling sites was used to account for possible effects of spatial autocorrelation. bimr analyses were run ten times for each area (using 1,000,000(burn in)/1,000,000(sampling) runs) and the gene flow values were obtained from the best run (which was selected on the basis of Dassignment as suggested by Faubet et al. ; when more than one run presented the same Dassignment, posterior probability and the visual analyses of likelihood plot for each run were used to determine the best run). BIMr results were analyzed also on the basis of the alpha value recovered for each factor (or combination of factor): “the sign and the magnitude of the alpha's tell us about the direction and the strength of the environmental factors” .
Isolation of the clusters detected by Geneland was checked using a series of assignment tests conducted in GeneClass . We performed four self-assignment tests using data from 2008 and 2009, separately, for each area. Then two assignment tests were performed assigning 2009 individuals using 2008 individuals as reference. For all six tests the individuals were divided according to the grouping recovered by the genetic structure analysis in Geneland. All tests were performed using Rannala and Mountain  method; 10,000 simulations with Paetkau et al.  method and a type I error rate P<0.01.
Genetic diversity and effective population size analysis
No significant deviation from Hardy-Weinberg- or linkage equilibrium was found. The genetic diversity indexes were very similar across sampling sites with different management and across years. The Kalø area showed a significantly higher allelic richness compared to Bjerringbro area (Table S1). Effective population size did not show any difference across area, year or management type (Tables S1 and S2).
Genetic structure analysis
Geneland analysis performed on samples from both years detected five clusters in Kalø and eight in Bjerringbro. In both cases no migrants were identified and samples from 2008 clustered together with the corresponding ones from 2009. All the sampling sites that did not constitute a cluster of their own were grouped accordingly to the nearest hedgerow (Fig. 2).
The figure shows the genetic clusters for each area with each cluster being included in a solid line (black for Bjerringbro, white for Kalø).
The pairwise Fst analyses showed higher Fst values in Bjerringbro than in Kalø; despite the presence of significantly differentiated pairs, no isolated (group of) sampling sites were found (Fig. 3). However, in Kalø there were two more distinct groups: KO1-KO2-KO3-KO4 and KO7-KC1-KC2-KC3 (Fig. 3) which partly agreed with the results from Geneland (Fig. 2) as the sampling sites in the two groups belonged to the same hedgerow (confirming the effect of fields as genetic barriers).
Gene flow analysis
BIMr analyses of Bjerringbro data converged in nine out of ten cases. “Hedgerow of belonging” was the most important factor in three out of nine runs. Further, in two single runs “road side” and “field of belonging”, respectively, were the most important explanatory factors. In other three runs “distance” (two runs) and “distance plus road side” (one run) were most important. When two factors were combined, “road side” and “field of belonging” were the least important ones (lowest absolute value of alpha). One run found the best fit to be the null model. The highest posterior probability and highest alpha were given to “hedgerow of belonging” alone. For the best run, the most important factor was “hedgerow of belonging” and the gene flow was higher for pairs of sampling sites belonging to the same hedgerow (see alpha values in Table 2).
All ten runs regarding Kalø samples converged. In nine out of ten cases “management type” was the most important factor alone (four runs), together with “hedgerow of belonging” (4 runs) or together with” field of belonging” (1 run). In cases where two factors were present, “management type” was the most important one (highest absolute number of alpha). The best run found “management type” alone to be the most important factor (with a posterior probability of 0.582, see Table 3) and, as the correlation coefficient (alpha) was negative, the migration was lowest from conventional to organic sampling sites and highest from organic to conventional ones (see Materials and Methods for details).
The self-assignment tests, performed on the samples from 2008 and 2009 for the two areas, were in good agreement with the population structure results as the percentage of mis-assigned individuals was generally low (Tables S3 and S4). The assignment test (2008 samples as reference for 2009) did not give clear cut results as most individuals could be assigned to more than one cluster. This was probably due to a lack of statistical power.
The main objectives of this study were: I) to determine the gene flow and the population structure patterns of B. lampros in agricultural areas with respect to landscape features (particularly: fields, hedgerows and roads) and II) to determine the effect of different land management strategies (specifically, pesticide use and agricultural intensiveness) on such patterns.
Genetic diversity and effective population size analysis
Genetic diversity indexes and effective population sizes were similar across all sampling sites both within and across areas (with the exception of allelic richness which was much higher in Kalø). There was also no indication of an effect of management. The reason for this homogeneity might lie in the fact that the populations are not genetically isolated (as proven by the fact that the majority of pairs, in the Fst analysis, did not differ significantly from each other).
Genetic structure analysis
The genetic structure analysis performed in Geneland showed that fields act as barriers to gene flow. The Geneland algorithm was chosen as it has been shown to be particularly good at finding genetic boundaries, especially when the population subdivision is recent . Many sampling sites constituted a cluster of their own, while the ones that were grouped, always belonged to the same hedgerow. Management did not seem to affect the population structuring. It should also be noted that samples from the two separate years always clustered together showing that genetic structure is consistent through years.
The patterns recovered by the Fst analysis were very different in the two areas (Fig. 4). The Fst values were higher in Bjerringbro than in Kalø; however, there were very few significantly different pairs in Bjerringbro compared to Kalø. In Kalø the Fst pattern is in good agreement with Geneland results as sampling sites lying on different field's sides are more differentiated than the ones sharing the same hedgerow. This results suggest a higher gene flow (lower Fst values), but an older isolation of sampling sites (more significantly different pairs) in Kalø compared to Bjerringbro. This difference might be due to a more consistent field structure in Kalø, where field management changed less over the years (Beate Strandberg, pers. comm.).
Gene flow analysis
The analysis of gene flow confirmed the results of the Fst analysis, showing a much higher gene flow in the Kalø area compared to Bjerringbro (Fig. 3). The barrier effect of fields was also confirmed and hedgerows were shown to be the main dispersal path in both areas (i.e. the highest gene flow occurred between sampling sites belonging to the same hedgerow, Fig. 4). A study by Eyre et al.  has shown the importance of vegetation cover and crop type on the activity of B. lampros and, despite not dealing directly with dispersal patterns, activity in the field was found not to be correlated with activity in the fields' margin. However, in the Kalø area there was a very clear effect of management on the strength and direction of gene flow. In particular, the highest gene flow went from organic towards conventional fields (Fig. 4 right, last three columns). The fact that this result was not found for Bjerringbro area suggests that the factor that makes a difference is the degree of intensity of the management and that only a low intensity of cultivation allows organic fields to act as genetic reservoirs. The positive effect of low-intensity cultivation has been shown for many wild species such as: corn buntings  and farmland birds in general , European hare  and for a large number of arthropods . Together with the previously mentioned studies, our work confirms, from a genetic point of view, the importance of low-intensity cultivations for the conservation of biodiversity. Moreover we show that, while organic fields can act as reservoirs for the re-colonization of conventional fields, the absence of pesticide use is not enough for a field to become a fauna-source for more managed areas. Therefore, our study discloses that (at least for this species) pesticide use and agricultural intensity are interdependent variables that should not be evaluated separately. It is however not possible to infer from these results the long-term effect of pesticide use and agricultural intensity due to the limited time-scale of the study (two years).
The results of self-assignment tests showed a good agreement with the genetic structure results as the number of mis-assigned individuals was generally low for each cluster. The results from Bjerringbro gene flow analyses showed that BO5 and BC1 are the most important source of emigrants; this is in accordance with the assignment tests from 2008 in which a lot of individuals from other clusters were assigned to the cluster containing these two populations. The results from Kalø for 2009 are also in agreement with the gene flow results as a lot of samples from other clusters were assigned to the clusters containing the conventional fields.
All the presented results suggest that tree-covered hedgerows are the preferred dispersal route used by B. lampros. On the other hand, fields represent a barrier to gene flow, suggesting a tendency for individuals to return to the same hedgerow. However, low-intensity organic fields can be used for dispersal, probably due to a lower death rate caused by mechanical treatments and a higher concentration of aphids. This finding confirms the hypothesis of Huusela-Veistola  which suggested that the increase in B. lampros abundance following a sharp decrease after spraying was due to immigration from neighboring organic fields. Going beyond this, our results also confirm the importance of intensity of cultivation, and in particular of a low yield cultivation with high weed abundance , . Future longer term and geographically broader studies might be able to determine with more accuracy the extent of the effect of each single variable.
Given the importance of B. lampros as a pest control agent, the results of this study can be important for the development of management strategies that maximize its beneficial effects . Moreover, the fact that tree-covered hedgerows and low-intensity cultivation were found to be fundamental for the gene flow in agricultural areas strengthen the importance of including landscape features and cultivation intensity as parameters for the evaluation of land management strategies.
Genetic diversity indexes for each sampling site and year. First column represent sampling site, (BO = Bjerringbro organic, BC = Bjerringbro conventional, KO = Kalø organic, KC = Kalø conventional) second column the year of sampling, third to fifth columns are the expected heterozigosity (He) allelic richness (Ar) and effective population size respectively (Ne) (see Materials and methods for details about their calculation).
Effective population sizes for all sampling sites calculated with moment based method. BO = Bjerringbro organic, BC = Bjerringbro conventional, KO = Kalø organic, KC = Kalø conventional.
Results of self-assignment tests for Bjerringbro. C1: BO1, BO2, BO4; C2: BO3; C3: BO5, BO6, BC1, BC2; C4: BC3; C5: BC4; C6: BC5; C7: BC6, BC7, BC8; C8: BC9.
We would like to thank all the people involved in the REFUGIA project for helpful discussions. Special thanks go to Zdenek Gavor, who performed the morphological identification of the beetle species.
Conceived and designed the experiments: CM LWA VL. Performed the experiments: CM LWA. Analyzed the data: CM LWA VL. Contributed reagents/materials/analysis tools: CM LWA VL. Wrote the paper: CM LWA VL.
- 1. Edwards CA, Sunderland K, George K (1979) Studies on polyphagous predators of cereal aphids. Journal of Applied Ecology 811–823.
- 2. Landis DA, Wratten SD, Gurr GM (2000) Habitat management to conserve natural enemies of arthropod pests in agriculture. Annual Review of Entomology 45: 175–201.
- 3. Chiverton PA (1986) Predator density manipulation and its effects on populations of Rhopalosiphumpadi (Horn.: Aphididae) in spring barley. Annals of Applied Biology 109: 49–60.
- 4. Cameron KH, Leather SR (2012) How good are carabid beetles (Coleoptera, Carabidae) as indicators of invertebrate abundance and order richness? Biodiversity and Conservation 21: 763–779.
- 5. Petersen MK (1999) The timing of dispersal of the predatory beetles Bembidion lampros and Tachyporus hypnorum from hibernating sites into arable fields. Entomologia Experimentalis et Applicata 90: 221–224.
- 6. Wallin H (1989) Habitat selection, reproduction and survival of two small carabid species on arable land: a comparison between Trechus secalis and Bembidion lampros. Ecography 12: 193–200.
- 7. Pederson M, Pedersen L, Abildgaard K (1990) Annual and diurnal activity of some Tachyporus species (Coleoptera: Staphylinidae) in two spring barley fields and a hedge. Pedobiologia 34: 367–378.
- 8. Eyre MD, Labanowska-Bury D, Avayanos JG, White R, Leifert C (2009) Ground beetles (Coleoptera, Carabidae) in an intensively managed vegetable crop landscape in eastern England. Agriculture, Ecosystems & Environment 131: 340–346.
- 9. Jansen JP, Defrance T, Warnier AM (2010) Effects of organic-farming-compatible insecticides on four aphid natural enemy species. Pest management science 66: 650–656.
- 10. Purvis G, Fadl A (2002) The influence of cropping rotations and soil cultivation practice on the population ecology of carabids (Coleoptera: Carabidae) in arable land. Pedobiologia 46: 452–474.
- 11. Hassan S, Bigler F, Bogenschütz H, Boller E, Brun J, et al. (1991) Results of the fifth joint pesticide testing programme carried out by the IOBC/WPRS-Working Group “Pesticides and Beneficial Organisms”. BioControl 36: 55–67.
- 12. Hassan S, Bigler F, Bogenschütz H, Boller E, Brun J, et al. (1988) Results of the fourth joint pesticide testing programme carried out by the IOBC/WPRS-Working Group “Pesticides and Beneficial Organisms”. Journal of Applied Entomology 105: 321–329.
- 13. Huusela-Veistola E (1996) Effects of pesticide use and cultivation techniques on ground beetles (Col., Carabidae) in cereal fields. Annales Zoologici Fennici 33: 197–206.
- 14. Honek A, Jarosik V (2000) The role of crop density, seed and aphid presence in diversification of field communities of Carabidae (Coleoptera). European Journal of Entomology 97: 517–526.
- 15. Purvis G, Curry J (1984) The influence of weeds and farmyard manure on the activity of Carabidae and other ground-dwelling arthropods in a sugar beet crop. Journal of Applied Ecology 271–283.
- 16. Milligan B, Hoelzel A (1992) Plant DNA isolation. Molecular genetic analysis of populations: a practical approach 59–88.
- 17. Andersen LW, Fog K, Damgaard C (2004) Habitat fragmentation causes bottlenecks and inbreeding in the European tree frog (Hyla arborea). Proceedings of the Royal Society of London, Series B: Biological Sciences 271: 1293–1302.
- 18. Goudet J (1995) FSTAT (version 1.2): a computer program to calculate F-statistics. Journal of Heredity 86: 485.
- 19. Peakall R, Smouse PE (2006) GENALEX 6: genetic analysis in Excel. Population genetic software for teaching and research. Molecular Ecology Notes 6: 288–295.
- 20. Ovenden JR, Peel D, Street R, Courtney AJ, Hoyle SD, et al. (2007) The genetic effective and adult census size of an Australian population of tiger prawns (Penaeus esculentus). Molecular Ecology 16: 127–138.
- 21. Pudovkin A, Zaykin D, Hedgecock D (1996) On the potential for estimating the effective number of breeders from heterozygote-excess in progeny. Genetics 144: 383–387.
- 22. Waples RS (1989) A generalized approach for estimating effective population size from temporal changes in allele frequency. Genetics 121: 379–391.
- 23. Guillot G, Mortier F, Estoup A (2005) GENELAND: a computer package for landscape genetics. Molecular Ecology Notes 5: 712–715.
- 24. Excoffier L, Lischer HEL (2010) Arlequin suite ver 3.5: a new series of programs to perform population genetics analyses under Linux and Windows. Molecular Ecology Resources 10: 564–567.
- 25. Faubet P, Gaggiotti OE (2008) A new Bayesian method to identify the environmental factors that influence recent migration. Genetics 178: 1491–1504.
- 26. Faubet P, Waples RS, Gaggiotti OE (2007) Evaluating the performance of a multilocus Bayesian method for the estimation of migration rates. Molecular Ecology 16: 1149–1166.
- 27. Piry S, Alapetite A, Cornuet JM, Paetkau D, Baudouin L, et al. (2004) GENECLASS2: a software for genetic assignment and first-generation migrant detection. Journal of Heredity 95: 536–539.
- 28. Rannala B, Mountain JL (1997) Detecting immigration by using multilocus genotypes. Proceedings of the National Academy of Sciences USA 94: 9197.
- 29. Paetkau D, Slade R, Burden M, Estoup A (2004) Genetic assignment methods for the direct, real-time estimation of migration rate: a simulation-based exploration of accuracy and power. Molecular Ecology 13: 55–65.
- 30. Safner T, Miller MP, McRae BH, Fortin MJ, Manel S (2011) Comparison of bayesian clustering and edge detection methods for inferring boundaries in landscape genetics. International Journal of Molecular Sciences 12: 865–889.
- 31. Brickle NW, Harper DGC, Aebischer NJ, Cockayne SH (2000) Effects of agricultural intensification on the breeding success of corn buntings Miliaria calandra. Journal of Applied Ecology 37: 742–755.
- 32. Smith HG, Danhardt J, Lindstrom A, Rundlof M (2010) Consequences of organic farming and landscape heterogeneity for species richness and abundance of farmland birds. Oecologia 162: 1071–1079.
- 33. Smith RK, Vaughan Jennings N, Harris S (2005) A quantitative analysis of the abundance and demography of European hares Lepus europaeus in relation to habitat type, intensity of agriculture and climate. Mammal Review 35: 1–24.
- 34. Hendrickx F, Maelfait J-P, Van Wingerden W, Schweiger O, Speelmans M, et al. (2007) How landscape structure, land-use intensity and habitat diversity affect components of total arthropod diversity in agricultural landscapes. Journal of Applied Ecology 44: 340–351.