Portal de Periódicos da CAPES

Biodiversity and the mitigation of climate change through bioenergy: impacts of increased maize cultivation on farmland wildlife

2011; Wiley; Volume: 3; Issue: 6 Linguagem: Inglês

10.1111/j.1757-1707.2011.01104.x

1757-1707

JANA GEVERS, Toke T. Høye, Christopher John Topping, Michael Glemnitz, Boris Schröder,

Environmental Conservation and Management

GCB BioenergyVolume 3, Issue 6 p. 472-482 Open Access Biodiversity and the mitigation of climate change through bioenergy: impacts of increased maize cultivation on farmland wildlife JANA GEVERS, JANA GEVERS Institute of Earth and Environmental Sciences, University of Potsdam, Karl-Liebknecht-Str. 24-25, 14476 Potsdam, Germany Leibniz Centre for Agricultural Landscape Research (ZALF) e.V., Eberswalder Str. 84, 15374 Müncheberg, GermanySearch for more papers by this authorTOKE THOMAS HØYE, TOKE THOMAS HØYE Department of Wildlife Ecology and Biodiversity, Aarhus University, Grenåvej 14, DK-8410, DenmarkSearch for more papers by this authorCHRIS JOHN TOPPING, CHRIS JOHN TOPPING Department of Wildlife Ecology and Biodiversity, Aarhus University, Grenåvej 14, DK-8410, DenmarkSearch for more papers by this authorMICHAEL GLEMNITZ, MICHAEL GLEMNITZ Leibniz Centre for Agricultural Landscape Research (ZALF) e.V., Eberswalder Str. 84, 15374 Müncheberg, GermanySearch for more papers by this authorBORIS SCHRÖDER, BORIS SCHRÖDER Institute of Earth and Environmental Sciences, University of Potsdam, Karl-Liebknecht-Str. 24-25, 14476 Potsdam, Germany Leibniz Centre for Agricultural Landscape Research (ZALF) e.V., Eberswalder Str. 84, 15374 Müncheberg, GermanySearch for more papers by this author JANA GEVERS, JANA GEVERS Institute of Earth and Environmental Sciences, University of Potsdam, Karl-Liebknecht-Str. 24-25, 14476 Potsdam, Germany Leibniz Centre for Agricultural Landscape Research (ZALF) e.V., Eberswalder Str. 84, 15374 Müncheberg, GermanySearch for more papers by this authorTOKE THOMAS HØYE, TOKE THOMAS HØYE Department of Wildlife Ecology and Biodiversity, Aarhus University, Grenåvej 14, DK-8410, DenmarkSearch for more papers by this authorCHRIS JOHN TOPPING, CHRIS JOHN TOPPING Department of Wildlife Ecology and Biodiversity, Aarhus University, Grenåvej 14, DK-8410, DenmarkSearch for more papers by this authorMICHAEL GLEMNITZ, MICHAEL GLEMNITZ Leibniz Centre for Agricultural Landscape Research (ZALF) e.V., Eberswalder Str. 84, 15374 Müncheberg, GermanySearch for more papers by this authorBORIS SCHRÖDER, BORIS SCHRÖDER Institute of Earth and Environmental Sciences, University of Potsdam, Karl-Liebknecht-Str. 24-25, 14476 Potsdam, Germany Leibniz Centre for Agricultural Landscape Research (ZALF) e.V., Eberswalder Str. 84, 15374 Müncheberg, GermanySearch for more papers by this author First published: 17 April 2011 https://doi.org/10.1111/j.1757-1707.2011.01104.xCitations: 47 Jana Gevers, tel. +49 331 977 2518, fax +49 331 977 2068, e-mail: jgevers@uni-potsdam.de AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinked InRedditWechat Abstract The public promotion of renewable energies is expected to increase the number of biogas plants and stimulate energy crops cultivation (e.g. maize) in Germany. In order to assess the indirect effects of the resulting land-use changes on biodiversity, we developed six land-use scenarios and simulated the responses of six farmland wildlife species with the spatially explicit agent-based model system ALMaSS. The scenarios differed in composition and spatial configuration of arable crops. We implemented scenarios where maize for energy production replaced 15% and 30% of the area covered by other cash crops. Biogas maize farms were either randomly distributed or located within small or large aggregation clusters. The animal species investigated were skylark (Alauda arvensis), grey partridge (Perdix perdix), European brown hare (Lepus europaeus), field vole (Microtus agrestis), a linyphiid spider (Erigone atra) and a carabid beetle (Bembidion lampros). The changes in crop composition had a negative effect on the population sizes of skylark, partridge and hare and a positive effect on the population sizes of spider and beetle and no effect on the population size of vole. An aggregated cultivation of maize amplified these effects for skylark. Species responses to changes in the crop composition were consistent across three differently structured landscapes. Our work suggests that with the compliance to some recommendations, negative effects of biogas-related land-use change on the populations of the six representative farmland species can largely be avoided. Introduction Renewable energies offer environmental benefits by reducing carbon dioxide (CO2) emissions through replacement of fossil fuels (Hjort-Gregersen et al., 2007). Hence, the EU regards the promotion of renewable energies as one way to fulfill its commitments to reducing CO2 emissions. One possibility to retain renewable energy is the production of biogas by anaerobic digestion of agricultural waste (e.g. manure), organic waste and agricultural crops in biogas plants. This results in biogas which consists mainly of CH4 and CO2 and can be used for the production of electricity and heat. Since the fermentation of organic manure is generally unprofitable and organic waste is of limited availability, the energy crop biogas is expected to become increasingly important (Holm-Nielsen et al., 2006; EEA, 2007). Because biomass, when produced by sustainable means, emits roughly the same amount of carbon during conversion to energy as is taken up during plant growth (Mc Kendry, 2001) it could bring substantial reductions in green house gases, compared with fossil fuels (Abraham et al., 2007). The choice of the energy crop is economically crucial for the farmer because a high biogas yield is required to offset high production costs (Walla & Schneeberger, 2008). Maize is considered to have the highest total biogas yield potential of field crops grown in central Europe (Amon et al., 2007) and will probably become increasingly important for energy production (Nielsen & Oleskowicz-Popiel, 2008). About 50% of the European landscape is used for agriculture (EEA, 2007) and is therefore strongly influenced by farmers' management decisions. These decisions are affected by farming opportunities and constraints and can lead to enormous changes in agro-ecosystems (Vandermeer & Perfecto, 1995). The production of biomass for bioenergy provides entirely new farming options since bioenergy cropping is targeting lower standards for product quality as food production (Faaij, 2006). Agricultural biomass production for energy purposes will, however, lead to changes in the composition and the spatial configuration of crops and could increase the pressure on farmland biodiversity (Tilman et al., 2001; EEA, 2007). Owing to bonus payments for providing energy generated from renewable resources [Erneuerbare-Energien-Gesetz (EEG) 2009], Germany is one of Europe's leaders in the production of biogas from energy crops (Abraham et al., 2007). This is likely to result in more biogas plants processing manure and maize together (Ammermann, 2007; Schöne, 2007), and maize for biogas production will probably be grown primarily near the biogas plants to minimize transport costs (Walla & Schneeberger, 2008). There are many environmental concerns regarding renewable energy, and numerous assessments of their possible environmental impacts have been conducted (e.g. Petersen, 2008; Eggers et al., 2009). However, large knowledge gaps exist with respect to the consequences of such land-use changes induced by biogas-related increase in maize cultivation for farmland animal populations (Dziewiaty & Bernardy, 2007). As every species has specific habitat requirements, detailed information on the species level is needed to assess the impact of land-use changes on community composition (e.g. Schröder et al., 2008). Furthermore, in order to evaluate consequences for the ecosystem, species from different trophic levels and with different life histories as well as ecological functions need to be taken into account (Jepsen et al., 2005). The following species have life-history strategies typically found in species inhabiting Europe's agro-ecosystems: skylark (Alauda arvensis), grey partridge (Perdix perdix), European brown hare (Lepus capensis syn. L. europaeus), field vole (Microtus agrestis), the spider species (Erigone atra) and the carabid beetle (Bembidion lampros). Several studies have been published evaluating habitat suitability of agricultural crops for one or more of these six species (Tapper & Barnes, 1986; Jenny, 1990; Pusenius & Viitala, 1993; Wakeham-Dawson et al., 1998; Downie et al., 2000; Purvis & Fadl, 2002; Irmler, 2003; Bro et al., 2004; Kuijper et al., 2009). However, there is little information available about their performance in maize fields, particularly in the context of heterogeneous agricultural landscapes. Furthermore, since the land requirements for bioenergy crops only started to increase in the last few years, the impact of the changes involved in the composition and spatial configuration of arable crops has not been adequately addressed until now (Dziewiaty & Bernardy, 2007). This study aims at evaluating the consequences of additional maize cultivation for biogas production and its spatial aggregation in a German agricultural landscape for the six characteristic farmland animal species listed above. There are various conflicts related to the expected land-use changes that could be important for some of these species. The influence of additional maize in the crop rotation will depend on the species-specific performance in this crop, which in turn depends partly on the sensitivity to maize-specific management events and their timing. Furthermore, the additional maize leads to a decrease in other crops (including rotational set-aside) and therefore to a decline in habitat diversity, which might also influence the species. We expect that these changes in crop composition will lead to lower abundances for species requiring high crop diversity, no changes for species not related to arable habitat including rotational set-aside and higher abundances for species that normally suffer from insecticides. The changes in the spatial crop configuration lead to an even more reduced crop diversity within the spatial aggregation of maize cultivation. We therefore expect the spatial concentration of maize to exacerbate species' responses. At the landscape scale, it is practically and economically almost impossible to assess the influence of land-use changes on populations of different animal species solely by field studies, especially because these might not only be determined by the area covered with certain crops, but also by the crops' spatial arrangement. However, a spatially explicit simulation model environment can be manipulated to control environmental factors, can be replicated, and circumvents the time constraints almost inevitably posed on field studies (Grimm et al., 2005). Therefore, simulation models are frequently applied in landscape and species management (Cousins et al., 2003; Rudner et al., 2007; Schröder et al., 2008; Ebrahimi et al., 2010). In this study, we used the thoroughly tested, spatially explicit agent-based Animal Landscape and Man Simulation System (ALMaSS) (Topping et al., 2003) capable of predicting population level responses to changes in landscape composition and configuration of six target species for different biogas related land-use change scenarios in three differently structured landscapes. Material and methods Description of the landscape model We used the ALMaSS, which combines agent-based animal models with a comprehensive and dynamic landscape simulation (Topping et al., 2003). The model was constructed for investigating the effect of changes in landscape structure and management on the abundance and distribution of different animal species. By simulating landscape processes related to land use, management decisions and vegetation growth, the landscape model in ALMaSS provides the environmental conditions used by the animal models and the facility to generate a diverse range of different scenarios. The landscape simulation model in ALMaSS is explicit in space and time. A detailed map of land cover serves as the spatial basis (Jepsen, 2004). The model is grid based, partitioned in 10 000 × 10 000 units and can therefore show detailed structures of 1 m2 resolution in a 100 km2 landscape (Topping et al., 2003). Each unit of the grid is allocated to a polygon which represents a structure in the landscape. The polygons are classified into one of several land cover classes (e.g. arable field, deciduous forest, natural grassland, cf. Topping et al., 2003). Every class has its own attributes and behaviours which are modelled by vegetation growth curves and a farm management simulator. Both, vegetation growth curves and farm management simulator, take inputs from a weather data file with daily records of mean temperature, mean wind speed and total daily precipitation. On arable fields (i.e. polygons that belong to the type 'arable fields'), the farm management simulator provides information about all farming activities that influence the model animals. All arable fields are assigned to a farm, which is classified into one of several predefined farm types (e.g. 'cattle farm'). Crop rotations and farm management decisions are specific to the farm type. The crop rotation is specified in a list containing all cultivated crops in the sequence they are supposed to be cultivated in. These lists typically contain 100 elements and multiple entries of each crop type. At the beginning of each simulation run, a random crop in the list is picked for each field. In the subsequent years, the cultivation of the crops in the list is simulated in a consecutive order. In this way, the proportion of a particular crop type in the list is translated into proportional cover of the crop on the cultivated area of the landscape. Specific management (e.g. sowing, pesticide application, harvest) is described for each crop and applied when predefined criteria are met (e.g. certain day length or thermal heat sum accumulation reaching a certain level, cf. Topping et al., 2003). The landscape model in ALMaSS was constructed for Northern Europe and is based on weather data and observed farm management practices from Denmark. Based on similarities in climatic zone and agricultural practices, we assume that the model could be transferred to a region in North East Germany. We used landscape maps from three different locations in Denmark (Fig. 1). Essentially, these landscapes represent arbitrary examples of the proportion and spatial configuration of land cover types (Table 1). Using three different landscapes allowed us to assess the general validity of our results concerning different landscape structures. The model landscape A is an arable landscape with a high proportion of grassland, forest and urban area. Landscape B holds the most arable land and marginal structures and the least forest. The landscape C contains the least grassland (Table 1). Figure 1Open in figure viewerPowerPoint The three 10 km × 10 km model landscapes: areas close to the towns of (a) Bjerringbro, (b) Herning and (c) Præstø. Table 1. The percentage cover of each landscape element in the three model landscapes (A, B and C) Landscape element Landscape A B C Arable land 52.1 70.5 66.1 Edge structures 0.5 2.3 0.8 Forest 21.8 11.1 20.5 Fresh water 1.9 0.6 0.7 Grassland 13.4 9.3 3.7 Housing area 8.1 4.3 5.2 Roads 2.2 1.8 3 Description of species models The agent-based animal models are built upon a 'state/transition' concept (Topping et al., 2003). Individuals develop through a series of different stages (e.g. eggs, juveniles) each with an associated range of behavioural patterns. If specific conditions are fulfilled, a transition into another state occurs. These conditions are characterized either by probabilities (e.g. overwintering probability) or by internal or external events (e.g. giving birth or being eaten) (Jepsen et al., 2005). In the current study, the time step of the model is one day for all species. The six species models simulate populations of the skylark (Topping & Odderskær, 2004), the grey partridge (Topping et al., 2010a), the European brown hare (Topping et al., 2010b), the field vole (Dalkvist et al., 2009), a sheet web spider (E. atra) (Thorbek & Topping, 2005) and a carabid beetle (B. lampros) (Bilde & Topping, 2004). We chose these six animals because of their different trophic levels in the food web, ecological properties, habitat requirements and behavioural strategies. Land-use scenarios We developed land-use scenarios of future crop production in North East Germany by using crop rotations based on land-use data and on transition probabilities from the Quillow catchment in North East Germany from 2003 (Glemnitz et al., 2011). The percentage of each crop in the rotation is presented in Table 2. To simulate the effects of the land-use changes on farmland animals we developed six scenarios. The scenarios differ in two aspects: (1) crop composition, which is determined by the additional cultivation of energy maize and (2) crop configuration, which is determined by the spatial structure of the energy maize cultivation (Table 3). In order to evaluate the species' responses to changes in crop composition, we established two scenarios with new crop rotations containing maize for biogas production. In the resulting 15Random and 30Random scenarios, the maize for energy production replaced 15% and 30% of other cash crops in the crop rotation (Holm-Nielsen et al., 2006; Dziewiaty & Bernardy, 2007). The new crop rotations contained 13% and 25% energy maize which corresponded to 15% and 30% of the cash crops (Tables 2 and 4). The three crop rotations (baseline, 15Random and 30Random) do not only differ in their percentage of maize, but as well in the overall percentage of the crops and rotational set-aside replaced by maize (Table 2). The replacement decisions were based upon medium-term financial and contractual obligations and the current situation on the world market; we assumed fixed proportions of certain cash crops and fodder crops, integrated the current relative economic attractiveness of cash crops on the world market and agricultural restrictions on crop rotations. Specifically we assumed all fodder crops (silage maize, clover-grass silage, fodder beet, seed grass) and cash crops grown under medium-term contract (oats) to be unchangeable. In the 15Random scenario, triticale, spring barley and rotational set-aside were completely replaced by energy maize. The same changes were made in the 30Random scenario and in addition, a proportion of the winter cereals was replaced by maize. Permanent grassland areas have not been altered in the scenarios as most of them are either bound to cattle farming by contract or outside of the production area. Table 2. Agricultural land use under different crop composition scenarios Crop Baseline 15Random 30Random Clover grass silage 2 2 2 Fodder beet 4 4 4 Seed grass 1 1 1 Fodder maize 10 10 10 Energy maize 0 13 25 Oats 2 2 2 Rotational set-aside 5 0 0 Spring barley 1 0 0 Triticale 7 0 0 Winter barley 8 8 5 Winter rape 23 23 20 Winter rye 2 2 2 Winter wheat 35 35 29 All values are expressed as average percentage crop cover in the arable land. Table 3. The percentage increase in maize cover among cash crops relative to the baseline scenario and spatial structure in the six land-use scenarios Scenario Energy maize Spatial structure of maize 15Random 15% Random 15Agg_small 15% Small-scale aggregation 15Agg_large 15% Large-scale aggregation 30Random 30% Random 30Agg_small 30% Small-scale aggregation 30Agg_large 30% Large-scale aggregation Table 4. Percentage of cash crops in the crop rotation under different crop composition scenarios Crop Baseline 15Random 30Random Energy maize 0 15.7 30.1 Oats 2.4 2.4 2.4 Rotational set-aside 6 0 0 Spring barley 1.2 0 0 Triticale 8.4 0 0 Winter barley 9.6 9.6 6 Winter rape 27.7 27.7 24.1 Winter rye 2.4 2.4 2.4 Winter wheat 42.2 42.2 34.9 All values are expressed as average percentage crop cover in the arable land. Since transportation cost is a major constraint in energy cropping (Walla & Schneeberger, 2008), we created additional scenarios in order to assess the species' responses to changes in the spatial crop configuration resulting from differences in the spatial structure of maize cultivation. We simulated an aggregated cultivation of maize by adding clusters particularly for biogas-related land use to the landscape design (Fig. 2). We assumed that farmers keep about half of their farm area for conventional husbandry and that they benefit from biogas production only as an additional source of income. Therefore, half of the agricultural area was used for these clusters. As the size of biogas plants and the associated cultivation area can vary strongly, we considered two sizes for the spatial aggregation of maize: (1) several small clusters of 200 ha each compared with (2) one large cluster with a size between 2600 and 3500 ha (depending on the amount of agricultural area in the respective landscape). The clusters were distributed randomly in the landscape. In the aggregation scenarios (15Agg_small, 15Agg_large, 30Agg_small and 30Agg_large) all additional maize is cultivated in these clusters. To accomplish this, we generated one crop rotation for the area outside the clusters (conventional land use) and one for the area within the clusters (biogas-related land use). Crop rotations are listed in the Supporting Information: Table S1 and S2. To quantify the spatial dependency of the aggregation clusters, we created three replicates for each cluster type (Agg Rep). These replicates differ in the location of the clusters in the landscape. Figure 2Open in figure viewerPowerPoint Example of the aggregated cultivation of maize around biogas plants. (a) Small-scale aggregation, (b) large-scale aggregation. All rotational set-aside was removed from both the 15Random and the 30Random scenario. However, from existing knowledge of habitat preferences of partridge, hare and skylark we identified rotational set-aside as a key crop type for these species. To separate the effects of the increase in maize and the loss of rotational set-aside, we created additional crop rotations identical to the 15Random and 30Random scenario except that the amount of rotational set-aside was equal to the original amount of rotational set-aside (SA100) or 50% of the original amount of rotational set-aside (SA50). This was achieved by reducing the frequency of all other crops in proportion to their occurrence. We evaluated the responses to these four new scenarios (15Random SA50, 15Random SA100, 30Random SA50 and 30Random SA100) on the skylark, partridge and hare. Data analysis For each scenario, we carried out simulations in each of the three landscapes and with each of the six species models. Additionally, we conducted simulations of the three replicates of the aggregations (Agg Rep). For all settings, we ran ten random replicates (Rand Rep) differing in initial conditions i.e. the distribution of individuals in the landscape and the first allocation of crops to the fields (Table 5). For each simulation, we recorded the number of adult females of each species monthly during 50 years. The models for larger vertebrates were run for 60 years before data was recorded while other species reached stable densities much more quickly and were only run for 10 years before data was recorded. Table 5. Simulations conducted in this study Scenario Simulation runs Total Baseline × 3 Landscapes × 6 Species × 5 Rand Rep 90 15Random × 3 Landscapes × 6 Species × 5 Rand Rep 90 15Agg_small × 3 Landscapes × 6 Species × 3 Agg Rep × 5 Rand Rep 270 15Agg_large × 3 Landscapes × 6 Species × 3 Agg Rep × 5 Rand Rep 270 30Random × 3 Landscapes × 6 Species × 5 Rand Rep 90 30Agg_small × 3 Landscapes × 6 Species × 3 Agg Rep × 5 Rand Rep 270 30Agg_large × 3 Landscapes × 6 Species × 3 Agg Rep × 5 Rand Rep 270 Total 1350 For every scenario, we ran the specified number of simulations. To quantify the effect of biogas-related land-use changes on the populations of the six animal species, we used the average number of females present in the landscape as measure of abundance. This is based on the assumption that males are not limiting population size. The effect of the changes in the crop composition was expressed as percentage decrease or increase of the population size relative to the relevant baseline. The effect of differences in the spatial crop configuration was calculated as mean abundance in the 15Agg_small and 15Agg_large scenarios, relative to mean abundances in 15Random scenario; and analogously for the 30Agg scenarios. To analyze the responses to the scenarios, we considered the mean abundance over all three landscapes. The significance of differences in female abundances between the scenarios was tested by the Wilcoxon rank-sum statistic. For the purpose of comparing densities of individuals between baseline and scenarios and between landscapes, we averaged across time series regardless whether the population went extinct in a particular run. Data analysis was carried out within the free software environment r 2.9.2 (R Development Core Team, 2009) using the packages rgr (Garrett, 2007), exactranktests (Hothorn & Hornik, 2009) and sciplot (Morales, 2010). Results The variation in population abundances was very small among random replicates but differed significantly between the three model landscapes for all six farmland species (Fig. 3). For instance, the skylark went extinct in the peri-urban landscape A, and mean abundances were consequently low. In contrast, both hares and voles were most numerous in Landscape A. All other species were most numerous in Landscape B where skylarks, for instance, had a mean abundance of more than 40 females/km2 (Fig. 3). Figure 3Open in figure viewerPowerPoint Female abundances of the six species in the three model landscapes under the baseline scenario. Bars indicate the 95% confidence interval (Chambers et al., 1983) calculated for the five random replicates. We also found changes in crop composition as dictated by the 15Random and 30Random scenarios to have very diverse effects on the six model species (Fig. 4). Relative to the baseline, both the 15Random and 30Random scenarios had a significant negative effect on the most mobile species: skylark (15Random: −33.3%, P=0.002; 30Random: −30.6%, P=0.003), partridge (15Random: −86.5%, P<0.0001; 30Random: −86.7%, P<0.0001) and hare (15Random: −37.6%, P<0.0001; 30Random: −38.8%, P<0.0001). For all three species, there was a distinct difference in abundance between the baseline simulation and the scenarios, but almost none between the two scenarios. Figure 4Open in figure viewerPowerPoint Effect of the evaluated scenarios. Female abundance under the 15Random and 30Random scenarios relative to abundance under the baseline scenario. Bars indicate the 95% confidence interval calculated for the differences between the three landscapes. Stars specify the level of significance, i.e. ***P≤0.001, **P≤0.01, *P≤0.05. The four reference scenarios developed to test the effect of rotational set-aside on the skylark, partridge and hare (i.e. 15Random SA50, 15Random SA100, 30Random SA50 and 30Random SA100) revealed strong responses to rotational set-aside in all three species. They show that the negative response in the skylark population to the gradual reduction of the proportion of rotational set-aside was more or less linear, whereas in both the hare and the partridge, scenarios with half the original proportion of rotational set-aside (15Random SA50 or 30Random SA100) resulted in abundances that were close to the result of running the corresponding scenario with no rotational set-aside (15Random or 30Random) (Table 6). In other words, removing half of the rotational set-aside resulted in more than half the reduction in abundance in the partridge and the hare, but not in the skylark. Table 6. Effect of the four reference scenarios (15Random SA50, 15Random SA100, 30Random SA50 and 30Random SA100) and the two main scenarios with random spatial allocation of crops (15Random and 30Random) Scenario Relative change in female abundance (%) Skylark Partridge Hare 15Random −26.5 −85.1 −36.3 15Random SA50 −15.4 −70.6 −33.9 15Random SA100 −7.1 −6.1 −8.8 30Random −28.3 −85.3 −39.3 30Random SA50 −16.2 −71.7 −34.3 30Random SA100 −7.9 −6.6 −12.9 The numbers are female abundance under the scenarios relative to abundance under the baseline scenario. The 15Random and 30Random scenarios had a significant positive effect on the spider (15Random: 8.0%, P<0.0001; 30Random: 16.2%, P<0.0001) and on the beetle (15Random: 2.7%, P=0.01; 30Random: 8.7%, P<0.0001). The changes in vole population abundance were not significant. Vole responses were below 1% under the 15Random and 30Random scenarios relative to the baseline simulation. Although the abundance of all six species differed between the three model landscapes, the responses to th