Plant diversity partitioning in Mediterranean ... - Wiley Online Library

Ecological Applications, 22(3), 2012, pp. 972–981 Ó 2012 by the Ecological Society of America

Plant diversity partitioning in Mediterranean croplands: effects of farming intensity, field edge, and landscape context ELENA D. CONCEPCIOŃ,1,3 FEDERICO FERNAŃDEZ-GONZA´LEZ,2 2

AND

MARIO DIÁZ1

1 Museo Nacional de Ciencias Naturales, CSIC, C/Serrano 115 bis, E-28006 Madrid, Spain Departamento de Ciencias Ambientales, Facultad de Ciencias Ambientales y Bioquı´mica, Universidad de Castilla-La Mancha, E-45071 Toledo, Spain

Abstract. Farmland biodiversity is affected by factors acting at various spatial scales. However, most studies to date have focused on the field or farm scales that only account for local (a) diversity, and these may underestimate the contribution of other diversity components (b diversity) to total (c) farmland diversity. In this work, we aimed to identify the most suitable management options and the scale at which they should be implemented to maximize benefits for diversity. We used a multi-scale additive partitioning approach, with data on plant diversity from 640 plots in 32 cereal crop fields from three agricultural regions of central Spain that differed in landscape configuration. We analyzed the relative contribution to overall plant diversity of different diversity components at various spatial scales and how these diversity components responded to a set of local (application of agri-environment schemes [AES] and position within the field) and landscape (field size and landscape connectivity and composition) factors. Differences in species composition among regions and then among fields within regions contributed most to overall plant diversity. Positive edge effects were found on all diversity components at both the field- and regional scales, whereas application of AES benefited all diversity components only at the field scale. Landscape factors had strong influences on plant diversity, especially length of seminatural boundaries, which increased species richness at both the field and the regional scales. In addition, positive effects of percentage of nonproductive land-uses in the landscape were found on all diversity components at the regional scale. Results showed that components that contributed most to overall plant diversity were not benefited by current AES. We conclude that agri-environmental policies should incorporate and prioritize measures aimed at the maintenance of seminatural boundaries and patches of nonproductive habitats within agricultural landscapes, through landscape planning, cross-compliance, or high nature-value farmland programs. These options will help to conserve overall plant diversity at regional scales, as well as the spillover of plant species from such seminatural elements into crops, especially in Mediterranean areas that still harbor extensive farming and relatively complex landscapes. Key words: agri-environment schemes; alpha diversity; beta diversity; edge effects; gamma diversity; landscape composition; landscape connectivity; regional effects.

INTRODUCTION There is an urgent need to identify the appropriate spatial scales at which conservation efforts should be implemented in order to counteract biodiversity loss in agricultural landscapes (Gabriel et al. 2006, 2010). Agrienvironment schemes (AES) of the European Union Common Agricultural Policy, and their North American and Australian relatives, are considered to be the main policy tool for reversing negative effects of agricultural intensification on biodiversity (Kleijn et al. 2011). However, the suitability of AES to reach this goal has been questioned because of their local application (i.e., field or farm scales), whereas agricultural intensification occurs at a wider range of spatial scales, from plots to Manuscript received 12 August 2011; revised 22 November 2011; accepted 7 December 2011. Corresponding Editor: V. C. Radeloff. 3 E-mail: [email protected] 972

landscapes (Benton et al. 2003, Concepcioń et al. 2008, 2012). In fact, effects of local agri-environmental management on biodiversity have been found to be constrained by factors such as landscape configuration (Tscharntke et al. 2005, Concepcioń et al. 2008, 2012, Bata´ry et al. 2011), or changes in land-use patterns within regions (Lo´pez-Jamar et al. 2011). Moreover, local biodiversity has been found to be benefited by AES application in fields and farms, but also by the proportion of land under agri-environmental management in the surrounding landscape (Gabriel et al. 2010). Thereby, an increasing number of studies highlight the need for adopting a multi-scale approach in both the design and the evaluation of AES in order to effectively conserve farmland biodiversity (Whittingham 2007, Concepcioń et al. 2008, 2012, Aviron et al. 2009, Gabriel et al. 2010). Despite these claims, most studies to date have focused on analyzing effects of either agricultural intensification or agri-environmental policies at the field or farm scales (but see Roschewitz et al. 2005, Gabriel et

April 2012

PLANT DIVERSITY PARTITIONING

al. 2006, Clough et al. 2007, Holzschuh et al. 2007, Flohre et al. 2011). These studies, that only account for a portion of farmland biodiversity, mostly local (a) diversity, may underestimate the contribution of other diversity components (b diversity) to total (c) farmland diversity, leading to erroneous assumptions and conclusions (Gabriel et al. 2006, Clough et al. 2007, Hendrickx et al. 2007). In the other side, multi-scale approaches based on the partitioning of the different components of total farmland diversity may help to identify proper scales to concentrate conservation efforts (Gering et al. 2003). The additive partitioning method (Allan 1975, Lande 1996) enables the division of total (c) diversity found in a given site (pooled number of species found in all sampling units within a site) into within-site (a) diversity (the mean number of species found per sampling unit within a site) and between-site (b) diversity (the difference between c diversity and a diversity, which measures the variation in species composition between sampling units within a site). In this way, c diversity at a given spatial scale equals to a- plus b-diversities. Moreover, additive partitioning can be applied at several spatial scales, since mean c diversity at a given scale is equal to a diversity at the next upper scale, so that total (c) diversity at one scale results from the a diversity of the lowest scale plus b diversities of intermediate scales. This decomposition allows evaluating the relative contribution of different spatial components to total diversity and, consequently, to identify the spatial scale at which environmental pressures or response schemes most influence species richness. In addition to scale issues, most evaluations of AES effectiveness have been carried out in Western and Central Europe, where agricultural intensity levels are considerably high, so that results obtained are difficult to extrapolate to more extensive, species-rich areas, such as the Mediterranean region (Kleijn and Sutherland 2003). These areas still harbor traditional agricultural landscapes that maintain high biodiversity levels and important populations of many species of European conservation concern (Kleijn et al. 2011). However, biodiversity conservation in these areas is especially uncertain, as it is threatened by either intensification or abandonment of current extensive management (Stoate et al. 2001). The importance of supporting farmers for maintaining extensive land-uses and thereby complex agricultural landscapes with high biodiversity levels in these areas, has been recognized only recently (Concepcioń et al. 2008, 2012, Kleijn et al. 2011). Consequently, despite some studies on this subject in Mediterranean systems have been developed in the last years (see e.g., Concepcioń et al. 2008, Guerrero et al. 2010, Jose´-Marı´ a et al. 2010, Bassa et al. 2011), further investigation on the factors influencing different components of biodiversity is still needed in order to find out the most suitable conservation strategies in these areas. In this work we analyze how a set of within-field and landscape factors affected distinct components of plant

973

diversity at different spatial scales in Mediterranean cereal croplands of central Spain. We aimed to identify the most suitable conservation strategies across scales to maximize benefits for diversity. We focused on arable weeds because they constitute a group strongly affected by agricultural intensification (Billeter et al. 2008, Kleijn et al. 2009) and with a special relevance in agroecosystems due to their close link to cultivation and the dependence of other groups of organisms at higher trophic levels on them (Marshall et al. 2003). We followed a multi-scale additive partitioning approach (Gabriel et al. 2006) using data on plant species from 640 plots in 32 crop-fields from three agricultural regions with different landscape configurations. Plant diversity components (a, b, and c diversity) were partitioned at the micro-scale (plots within fields), the meso-scale (fields within regions), and the macro-scale (the three regions as a whole). We analyzed the relative contribution of different components at each spatial scale to overall plant diversity and how these diversity components responded to a set of within-field (application of AES and position within the field) and landscape factors (field size, length of seminatural boundaries, and percentage of nonproductive land uses in the landscape surrounding fields) at the micro- and meso-scales. To date, similar approaches have been applied in central Europe to evaluate the relative effects of AES with strong land use intensity limitations (mostly organic farming) on plant diversity in agro-ecosystems differing in landscape complexity (Roschewitz et al. 2005, Gabriel et al. 2006, Clough et al. 2007, Holzschuh et al. 2007). We applied this approach in a low-intensity farming system (Mediterranean dry-cereal croplands of central Spain) in which AES application involved a moderate reduction of land-use intensity as compared with conventional management. In addition, we explicitly incorporated to the analyses a complete set of landscape metrics describing the landscape configuration of the study sites, in order to discriminate which of them influenced different plant diversity components at each spatial scale. We expect that plant diversity components at higher spatial scales (i.e., meso- and macro-scales) will contribute more to overall diversity than local components (Roschewitz et al. 2005, Gabriel et al. 2006, Clough et al. 2007, Holzschuh et al. 2007, Flohre et al. 2011), even more when considering the high levels of extensification and landscape complexity of the study areas. We also expect that plant diversity components at higher spatial scales will be mostly affected by landscape configuration rather than by local management. If these expectations are met, the application of current AES will have a limited scope to enhance plant diversity at regional scales, so that further or alternative policy tools will be required to preserve farmland biodiversity. METHODS Sampling design Three agricultural regions of Castilla-La Mancha (central Spain), in which AES with a biodiversity

974

Ecological Applications Vol. 22, No. 3

ELENA D. CONCEPCIOŃ ET AL.

conservation objective had been applied for at least 5 years since 1995–1996 until 2003 (6.7 6 0.1 years on average [mean 6 SE]), were selected as study areas. The most widely implemented AES in Spain (Oñate et al. 1998), which were aimed at the conservation of biodiversity in extensive cereal crops and whose prescriptions basically consisted of reducing pesticide and fertilizer application, restricting farming practices for some dates and maintaining unplowed strips covering 3% of the fields, was selected (see Kleijn et al. [2006] and Concepcioń and Dı´ az [2011] for details). These prescriptions, which are basically the same as those included in AES currently applied in Spanish cereal croplands, were applied simultaneously in study fields, excepting for the maintenance of unplowed strips within fields, which was voluntary and had a very low level of uptake in study sites (E. D. Concepcioń, F. Fernańdez-Gonza´lez, and M. Dı´ az, unpublished data). Afterward, three regions differing in landscape configuration from among available areas with these AES were selected. The three areas chosen are mainly devoted to dry cereal cropping (70–85% area covered by arable fields), especially barley (Hordeum vulgare and H. distichon) and wheat (Triticum aestivum and T. durum). Study areas are also close to each other (60–80 km), but differ in soil types and landscape configuration: La Guardia (39847 0 N, 3839 0 W, 620–670 m above sea level), a simple coarse-grained agricultural landscape, with basic soils; Huecas (40800 0 N, 4812 0 W, 520–580 m above sea level), the intermediate landscape in which basic soils also predominate; and Retuerta del Bullaque, (39827 0 N, 4822 0 W, 710–750 m above sea level), the most complex landscape, with acidic soils. In each study area, seven pairs of arable fields, each pair with one field with AES and a control field farmed conventionally, were selected. Fields within pairs were devoted to the cultivation of the same cereal species and were selected close to each other, minimizing differences between them in size, shape, soil type, and landscape context. During spring 2003, the species richness of vascular plants was measured in study fields. Plant data were collected in ten plots of 5 3 1 m spaced 5 m apart placed along the inner edge of each field, adjacent to a field margin, and 10 more in the center, along a parallel transect at least 50 m away from any field edge (see Kleijn et al. [2006] for details). Sampling was carried out from mid-May to mid-June, one to two weeks before harvesting. Fields within pairs were surveyed by the same person on the same date. Plant data were finally collected in 32 fields, as five field pairs initially selected for the study had to be discarded because at least one of the fields within each of these pairs was harvested before plant sampling. A set of landscape metrics of study areas were measured using ArcView GIS 3.2 (ESRI 2000) in order to characterize landscape configuration around each study field. Specifically, size of focal fields (in ha), and length of seminatural boundaries (mostly grassy strips between fields and, occasionally, shrubby and riparian

vegetation) and percentage of nonproductive land uses in 500 m radius buffer areas centered on focal fields were recorded. We selected these metrics in order to account for different components of landscape that have been found to affect most local diversity in study sites: field configuration, landscape connectivity, and landscape composition (Concepcioń et al. 2008). Landscape metrics significantly differed among the three study areas, but did not vary between the fields within pairs, either within or among regions. In particular, the mean patch size increased and the mean number of patches, as well as the length of seminatural boundaries, decreased from the most complex to the simplest area (see Concepcioń et al. [2008] for details). Plant diversity data The spatial components of plant diversity were separated using an additive partitioning approach (Allan 1975, Lande 1996), similar to that applied by Gabriel et al. (2006). The total species richness of plants recorded in study sites were divided into diversity components (a, b, and c) at three spatial scales, micro- (plots within fields), meso- (fields within regions) and macro-scale (the three regions), for each combination of the two treatment factors (i.e., position in the field and farming system: edge/ with AES, edge/conventional, center/with AES, center/ conventional). The c diversity (total species diversity) was the number of species found in the pooled sampling units (i.e., in all plots per field, in all fields per region, or in all regions as a whole), the a diversity (within-site diversity) was the mean number of species found in the sampling unit considered at each scale (i.e., per plot in a field, per field in a region, or per region for the overall data), and the b diversity (between-site diversity) was the mean species turnover among sampling units (i.e., the difference between c and a diversity) at each spatial scale. Data analysis The relative contribution to overall plant diversity of the different diversity components at the three spatial scales (micro-, meso- and macro-) was calculated on the basis of their additive properties (c macro ¼ a micro þ b micro þ b meso þ b macro). Then, linear mixed-effects models (Pinheiro and Bates 2000), which account for nonindependent errors due to hierarchically nested sampling design, were used to test for the effects of a set of within-field and landscape-scale traits on the different components of plant diversity (a, b, and c diversity) at both the micro- and the meso-scale. Field pair was included as a random factor in mixed effects models to account for the paired sampling design. Position in the field (edge vs. center), farming system (with AES vs. conventional), region (only at the microscale, as at the meso-scale plant data were grouped by regions), size of study fields (in ha), length of seminatural boundaries (in m) and percentage of nonproductive (i.e., non-farmed) land-uses in the buffer-areas of 500 m-radius around focal fields were included as fixed explanatory

April 2012


parameters. First-order interactions among predictors were also included in models. At the meso-scale, landscape metrics were included in models averaged for each combination of categorical factors (i.e., farming system, position within the field and region). Best fitted models for each diversity component at each spatial scale were selected using the likelihood ratio test. First, nonsignificant interactions were removed and then explanatory parameters were dropped from models stepwise, computing the likelihood ratio and removing variables with nonsignificant values until all variables included in the model when dropped produced a significant decrease in the model likelihood. Nonsignificant predictors included in significant interactions were not removed from models (Crawley 2007). All statistical analyses were performed with R (R Development Core Team 2009). RESULTS A total of 292 species of plants were found in study sites (Supplement), of which 84% were found in the edges of fields with AES, 80% in the edges of conventional fields, 64% in the center of fields with AES, and 61% in the center of conventional fields. Regarding diversity components (Fig. 1), the highest contribution to overall species richness was made by b diversity at the macro- and meso-scales (up to 44% and 21%, respectively). Contribution of a and b diversity at the micro-scale was similar (up to 9% and 11%, respectively). Therefore, the highest contribution to total species richness was made by differences in species found (b diversity) among regions (macro-scale) and, secondly, among fields within regions (meso-scale). Results of mixed effects models (Table 1) showed that position in the field had the greatest effects on species richness, mostly at the micro-scale, albeit it also had strong effects at the meso-scale. All diversity components at the two spatial scales were higher in the field edges than in the centers of fields (Fig. 2). Application of AES only had significant positive effects on all diversity components at the micro-scale (Fig. 3; estimated effects 4.68 6 1.12 [mean 6 SE], t ¼ 4.18, P ¼ 0.0002, for a diversity; 3.46 6 1.45, t ¼ 2.38, P ¼ 0.02, for b diversity; 7.39 6 2.21, t ¼ 3.35, P ¼ 0.002, for c diversity). Study area was also among the parameters with greater influence on plant species richness at this scale (the only one at which it was examined). All diversity components increased from the simplest to the most complex region (Fig. 4). Regarding landscape factors, length of seminatural boundaries around focal fields had strong positive effects on c and a diversity and marginally affected b diversity at the micro-scale (Fig. 5a). At the meso-scale, strong effects of length of seminatural boundaries in the landscape were also found on c and b diversity (Fig. 5b, left hand). Percentage area of nonproductive land uses around focal fields had additional positive effects on all diversity components at the meso-scale (Fig. 5b, right hand). Moreover, size of focal fields interacted with position in the field at the meso-scale, with negative

975

FIG. 1. Components of plant diversity (mean a diversity at the micro-scale, and b diversities at the micro-, meso-, and macro-scales [þSE]) in the edges and the centers of fields with agri-environment schemes (AES) or under conventional management. No SE is provided for b diversity at the macro-scale because this parameter was measured only once for each combination of the classification factors, i.e. for the three study areas as a whole.

effect on a diversity only in the field edges (estimated effect ¼1.11 6 0.42, t ¼ 2.66, P ¼ 0.045). At the microscale, position in the field interacted with study area, reflected by larger differences in c and a diversity between the edges and the centers of fields in the complex area (Retuerta del Bullaque; estimated effects, 11.87 6 1.63, t ¼ 7.30, P , 0.0001, for a diversity; 22.25 6 3.55, t ¼ 6.27, P , 0.0001, for c diversity) compared with the intermediate (Huecas; estimated effects, 3.03 6 1.51, t ¼ 2.01, P ¼ 0.05, for a diversity; 9.14 6 3.29, t ¼ 2.78, P ¼ 0.008, for c diversity) and the simple region (La Guardia; estimated effects, 6.75 6 2.30, t ¼ 2.94, P ¼ 0.005, for a diversity; 15.67 6 5.02, t ¼ 3.12, P ¼ 0.003, for c diversity) areas. Study area also interacted with field size at the micro-scale, as a diversity increased with size of study fields only in the intermediate area (Huecas; estimated effect, 0.28 6 0.09, t ¼ 3.10, P ¼ 0.004). Last, at the meso-scale, effects of percentage of nonproductive land around focal fields interacted with application of AES, which positively affected c and b diversity but only in fields without AES (estimated effect, 21.62 6 6.16, t ¼ 3.51, P ¼ 0.025, for b diversity; 33.76 6 9.77, t ¼ 3.46, P ¼ 0.026, for c diversity). DISCUSSION Through a novel hierarchical multi-scale approach, we assessed the relative contribution of factors related to agricultural management at different spatial scales to the conservation of overall plant diversity in extensive, species-rich Mediterranean cereal croplands. Our study sites harbored high levels of species richness of plants, typical of Mediterranean regions (Holzner and Immonen 1982), as compared with other European countries (see supplementary material in Kleijn et al. [2006]). The



976

TABLE 1. Best fitted linear mixed-effects models for the effects of position in the field (edge vs. center), farming system (agrienvironment schemes [AES] vs. conventional), study area (simple, medium, complex), and landscape complexity (size of focal field and length of seminatural boundaries and percentage of nonproductive land uses in 500-m buffers around focal fields) on the different components of plant diversity (a, b, and c diversity) at the micro- (plots within fields; n ¼ 64 plots) and the mesoscale (fields within regions; n ¼ 12 fields). Micro-scale diversity Effects

df

Farming system Field position Study area Field size (ha) Length of natural boundaries (m) Nonproductive area (%) Farming system 3 field position Farming system 3 study area Farming system 3 field size Farming system 3 length of natural boundaries Farming system 3 nonproductive area Field position 3 study area Field position 3 field size Field position 3 length of natural boundaries Field position 3 nonproductive area Study area 3 field size Study area 3 length of natural boundaries Study area 3 nonproductive area Model AICc

1 1 2 1 1

a (þ) 20.68*** (þ) 50.00*** (þ) 40.26*** 1.45 (þ) 14.31***

b

Meso-scale diversity c

(þ) 7.56* (þ) 33.07*** (þ) 19.41***

(þ) 15.16*** (þ) 49.38*** (þ) 33.59***

(þ) 3.53

(þ) 7.84**

1 1 2 1 1

a (þ) 45.91** — 3.42 (þ) 33.04**

1 2 1 1

b 0.24 (þ) 15.60* —

2.81 (þ) 22.52** —

(þ) 64.08**

(þ) 39.57**

(þ) 19.32*

(þ) 29.21**

11.19* 7.96**

c

7.92*

3.67* 7.09*

1 2 2

4.26*

2 375.64

416.03

459.66

86.36

99.99

111.06

Notes: Sign of effects (þ or ) and F values of parameters included in best fitted models for each response variable are given. Cells with dashes indicate that the parameter was not included in models from the beginning of the analysis because we have only three observations to test its effects. Empty cells indicated parameters that were initially considered in models but finally were not included in best fitted models according to log-likelihood tests. AICc is the Akaike information criterion adjusted for small sample sizes. Error df are, for the micro-scale tests, 40, 45, and 43 (for a, b, and c, respectively) and for the meso-scale tests, 5, 4, and 4, except for study area, for which error df is 13 in all cases. * P , 0.05; ** P , 0.01; *** P , 0.001; P , 0.10.

highest contribution to total plant diversity (Fig. 1) was accounted for by differences in species composition among regions (b diversity at the macro-scale) and secondly by differences among fields (b diversity at the meso-scale). Similar results have been found in previous studies (Gabriel et al. 2006, Clough et al. 2007, Flohre et al. 2011). The prominent regional role was possibly related to differences in species pools among regions (Clough et al. 2007) due to land-use histories or biogeographic traits, such as differences in soil types among study areas, with basophilic flora prevailing in La Guardia and Huecas, and acidophilic species in Retuerta del Bullaque. It must be noted that both local plant diversity and species pool richness are expected to be higher on calcareous soils in many temperate ecosystems of the northern hemisphere (Grime 2001, Pa¨rtel 2002, Rey-Benayas and Scheiner 2002, Ewald 2003). Therefore, the high species richness in Retuerta del Bullaque (the complex area, Fig. 4) may be attributable to effects of more extensive management in this region due to lower productivity of soils compared with the other regions. Similar results of higher floristic richness in arable fields of central Europe

on acidic soils related to less intensive agricultural management have been reported (Chytry´ et al. 2003). On the other hand, differences among fields within regions also accounted for a large proportion of total plant diversity, a fact that was most probably due to landscape heterogeneity (Clough et al. 2007). Most species were found in the edges of fields (Fig. 2), in both conventional fields and managed under AES, albeit that more species were found in edges of fields where AES were applied compared with edges of conventional ones (84% vs. 80% of total species found). A similar result was found by Jose´-Marı´ a et al. (2010) in dry cereal croplands of north-eastern Spain. In fact, position in the field (edge vs. center) had the strongest effects on all diversity components at the micro-scale (Table 1). At the meso-scale, position in the field also affected all diversity components, especially a diversity, but landscape traits had greater effects on either c or b diversity. These outstanding effects of position in the field can be explained because of an edge effect (Forman 1995) resulting from species migration from the adjacent boundaries and lower intensity of agricultural practices in the edges than in the centers of fields (Marshall 1989,

April 2012


977

FIG. 2. Box plots of plant diversity components (a, b, and c diversity) at (a) the micro-scale and (b) the meso-scale in the edges (E) and the centers (C) of study fields. Center lines represent medians, and the outer lines represent the inter-quartile range. Whisker lines represent the whole range of data that lie within one and a half times the interquartile range (1.5 3 IQR). Significant edge effects were found for all diversity components at both spatial scales (see also Table 1).

2009, Wilson and Aebischer 1995, Marshall and Moonen 2002). Additionally, at the micro-scale, differences in species richness between the edges and the centers of fields were larger in the most complex area, as field edges in this area reached higher diversity levels. This suggests a stronger edge effect in fine-grained and connected landscapes, which may be the result of additive edge effects of the whole network of field boundaries present in the landscape (Fernańdez et al. 2002). In contrast, farming system (application of AES) only affected plant diversity at the micro-scale (Fig. 3) and thus, although positive effects of AES were found for all diversity

components at this scale, it contributed considerably less to total plant diversity than edge effects. Landscape-scale factors were of great influence in all diversity components at either the micro- or the mesoscale. Length of seminatural boundaries in the landscape had strong positive effects on plant diversity at both the micro- and the meso-scales (Fig. 5), as these elements act as reservoirs and dispersal corridors from and by which more species reach crops (Marshall and Moonen 2002, Marshall 2009). Nonetheless, at the micro-scale, differences among regions resulted much more relevant. On this respect, regional differences in plant diversity are probably due to differences in soil types among regions,

978



FIG. 3. Box plots of plant diversity components (a, b, and c diversity) at the micro-scale in fields with AES (1) and conventional schemes (0). Whisker lines represent the whole range of data that lie within one and a half times the interquartile range (1.5 3 IQR). Application of AES significantly increased all diversity components at the micro-scale (see also Table 1).

though these effects might in some way be confused with effects caused by differences in landscape complexity among regions, as all diversity components increased from the simplest to the most complex region (Fig. 4). On the other hand, percentage of nonproductive land uses in the landscape had also strong effects on all the components of plant diversity at the meso-scale (Fig. 5b, right-hand panel), as these remnants of nonproductive habitats also act as refuge and source of organisms within the crop matrix in agricultural landscapes

(Benton et al. 2003). Therefore, the maintenance of seminatural boundaries in agricultural landscapes seems to be important to conserve plant diversity within fields. However, at regional scales, besides increasing landscape connectivity, it may be also necessary to maintain or restore patches of nonproductive habitats within the landscape, in order to enhance plant diversity. Additionally, landscape metrics interacted with other predictors in their effects on plant diversity. Size of focal fields negatively affected local diversity at the meso-scale,

FIG. 4. Box plot of plant diversity components (a, b, and c diversity) at the micro-scale for the three study areas (1, La Guardia, the simplest landscape; 2, Huecas, the intermediate landscape; and 3, Retuerta del Bullaque, the most complex landscape). Whisker lines represent data values that lie within one and a half times the interquartile range (1.5 3 IQR), and points represent outliers, i.e., data values above or below 1.5 3 IQR. All diversity components at the micro-scale increased from the simplest to the most complex area (see also Table 1).

April 2012


979

FIG. 5. Relationships between length of seminatural boundaries (m) and percentage of nonproductive land use in the 500-m radius buffer-areas around focal fields, and different components of plant diversity (a, open circles, dotted lines; b, gray circles, dashed lines; and c, black circles, solid lines) at (a) the micro-scale and (b) the meso-scale. Length of seminatural boundaries significantly increased all diversity components at the micro-scale, and b and c diversity at the meso-scale. Percentage of nonproductive land uses significantly increased all diversity components at the meso-scale (see also Table 1).

but only in the field edge, probably due to a lower perimeter : area ratio and then, reduced edge effects in larger fields. Field size and shape have been found to interact with edge effect giving rise to varying effects within fields because of additive effects of all field boundaries in a given point of the field that depend on the distance and the intensity of each punctual edge effect (Fernańdez et al. 2002). Thus, smaller fields would be more affected by edge effects because of shorter distances to all field boundaries. This fact would also explain why edge effects were larger in the complex, fine-grained, landscape (Retuerta del Bullaque), highlighting the need

to conserve the density of seminatural boundaries in agricultural landscapes. This is in agreement with results reported by Gabriel et al. (2005), who found that plant species in cereal crops were more affected by the entire net of edges in the landscape than by the adjacent field edge. Besides, effects of percentage of nonproductive land-uses around focal fields interacted with effects of application of AES, increasing plant diversity at the meso-scale but only in conventional fields, most likely because higher amounts of nonproductive habitats would compensate for the effects of more intensive local management in conventional fields. Such compensating effect of land-

980



scape context has been found to occur at high levels of landscape complexity (Concepcioń et al. 2008, 2011, Bata´ry et al. 2011), which may correspond to those of our study areas (Kleijn et al. 2006, Concepcioń et al. 2008). Overall, most contribution to total biodiversity was provided by differences in species composition among regions and secondly among fields. The first component may be attributable to differences in regional species pools, mostly caused by biogeographic traits (i.e., soil types), and the second, to landscape heterogeneity within regions. In fact, whereas plant diversity at the micro-scale (within fields) was affected by local and landscape traits, diversity at the meso-scale (within regions) was almost exclusively affected by landscape traits, with the only exception of position in the fields. Field position (edge vs. center) was among the factors of greatest influence in all diversity components at both the micro- and the meso-scale. Positive edge effects were additionally intensified by landscape complexity, which emphasizes the relevance of maintaining a dense network of seminatural boundaries in agricultural landscapes to support the dispersal of plant species along fringes and into the crop fields, and thereby the conservation of overall plant diversity. In contrast, benefits of AES for plant diversity were only significant at the micro-scale, and additionally their effects were compensated by the complexity of the surrounding landscape. Nonetheless, AES, despite acting solely at the field-scale, increased not only local diversity, but also the differences in species richness within fields, which slightly increased their relative contribution to total plant diversity. Regarding landscape factors, they strongly affected all diversity components, both within fields and within regions. Hence, preservation of landscape connectivity through maintaining or restoring seminatural boundaries in agricultural landscapes, in order to enhance plant diversity both at the field- and regional scales, should turn into a central issue in agrienvironmental policies. Furthermore, maintaining or restoring remnants of nonproductive habitats that can act as source of species in order to enhance plant diversity at regional scales should be additionally integrated into such conservation programs. Conclusions and management implications Our results highlight the need of taking into account a wider range of spatial scales in either the design or the evaluation of agri-environmental policies (Gabriel et al. 2010, Concepcioń et al. 2012). We have shown how components mostly contributing to overall plant diversity are not targeted in current agri-environmental policies, and that such components are not benefited by AES applied to date. Agri-environmental programs should incorporate and prioritize landscape management options in order to benefit components that contribute most to overall biodiversity. Moreover, local diversity will be also benefited by these initiatives. Hence, agri-environmental programs should adopt a

hierarchical multi-scale approach, which combines landscape management options with field-scale measures with specific conservation objectives (Concepcioń et al. 2011). In this new multi-scale strategy, focus has to be pointed at the conservation of seminatural boundaries and remnants of nonproductive habitats within agricultural landscapes. Such a focus is especially relevant in extensive and complex areas with high biodiversity levels, such as the Mediterranean region, in order to avoid landscape simplification and the subsequent loss of biodiversity through further intensification or land abandonment. This strategy can be implemented by means of compulsory measures within cross-compliance or agri-environmental programs defining the spatial distribution of fields or farms in which a variety of landscape management options, including the maintenance or the restitution of seminatural elements, can be applied. High nature value farmland programs can also include measures supporting farmers for maintaining current practices and farm configuration, which guarantee the conservation of complex agricultural landscapes with high biodiversity levels. ACKNOWLEDGMENTS We acknowledge J. de Esteban and R. A. Baquero for the field work, farm owners and J. A. Millań for their collaboration, and L. Barrios for the statistical help. Suggestions made by D. Gabriel and J. Marshall on a first draft highly contributed to improve it. Suggestions made by two anonymous referees were also very helpful. This work was partly funded by the EU Project QLK5-CT-2002-1495 ‘‘Evaluating Current European Agrienvironment Schemes to Quantify and Improve Nature Conservation Efforts in Agricultural Landscapes (EASY).’’ LITERATURE CITED Allan, J. D. 1975. Components of diversity. Oecologia 18:359–367. Aviron, S., H. Nitsch, P. Jeanneret, S. Buholzer, H. Luka, L. Pfiffner, S. Pozzi, B. Schu¨pbach, T. Walter, and F. Herzog. 2009. Ecological cross compliance promotes farmland biodiversity in Switzerland. Frontiers in Ecology and the Environment 7:247–252. Bassa, M., C. Boutin, L. Chamorro, and X. Sans. 2011. Effects of farming management and landscape heterogeneity on plant species composition of Mediterranean field boundaries. Agriculture, Ecosystems and Environment 141:455–460. Bata´ry, P., A. Ba´ldi, D. Kleijn, and T. Tscharntke. 2011. Landscape-moderated biodiversity effects of agri-environmental management—a meta-analysis. Proceedings of the Royal Society B 278:1894–1902. Benton, T. G., J. A. Vickery, and J. D. Wilson. 2003. Farmland biodiversity: is habitat heterogeneity the key? Trends in Ecology and Evolution 18:182–187. Billeter, R., et al. 2008. Indicators for biodiversity in agricultural landscapes: a pan-European study. Journal of Applied Ecology 45:141–150. Castroviejo, S., editor. 1986–2011. Flora ibe´rica. Consejo Superior de Investigaciones Cientı´ ficas, Madrid, Spain. Chytry´, M., L. Tichy´, and J. Rolecek. 2003. Local and regional patterns of species richness in Central European vegetation types along pH/calcium gradient. Folia Geobotanica 38:429– 442. Clough, Y., A. Holzschuh, D. Gabriel, T. Purtauf, D. Kleijn, A. Kruess, I. Steffan-Dewenter, and T. Tscharntke. 2007. Alpha and beta diversity of arthropods and plants in organically and conventionally managed wheat fields. Journal of Applied Ecology 44:804–812.

April 2012


Concepcioń, E. D., and M. Dı´ az. 2011. Field, landscape and regional effects of farmland management on specialist openland birds: does body size matter? Agriculture, Ecosystems and Environment 142:303–310. Concepcioń, E. D., M. Dı´ az, and R. A. Baquero. 2008. Effects of landscape complexity on the ecological effectiveness of agri-environment schemes. Landscape Ecology 23:135–148. Concepcioń, E. D., et al. 2012. Interactive effects of landscape context constrain the effectiveness of local agri-environmental management. Journal of Applied Ecology, in press. Crawley, M. J. 2007. The R book. John Wiley, Chichester, UK. ESRI. 2000. ArcView GIS 3.2a. Environmental Systems Research Institute, Redlands, California, USA. Ewald, J. 2003. The calcareous riddle: Why are there so many calciphilous species in the Central European flora? Folia Geobotanica 38:357–366. Fernańdez, C., F. J. Acosta, G. Abella´, F. Lo´pez, and M. Dı´ az. 2002. Complex edge effect fields as additive processes in patches of ecological systems. Ecological Modelling 149:273–283. Flohre, A., et al. 2011. Agricultural intensification and biodiversity partitioning in European landscapes comparing plants, carabids and birds. Ecological Applications 21:1772–1781. Forman, R. T. T. 1995. Land mosaics: the ecology of landscapes and regions. Cambridge University Press, Cambridge, UK. Gabriel, D., I. Roschewitz, T. Tscharntke, and C. Thies. 2006. Beta diversity at different spatial scales: plant communities in organic and conventional agriculture. Ecological Applications 16:2011–2021. Gabriel, D., S. M. Sait, J. A. Hodgson, U. Schmutz, W. E. Kunin, and T. G. Benton. 2010. Scale matters: the impact of organic farming on biodiversity at different spatial scales. Ecology Letters 13:858–869. Gabriel, D., C. Thies, and T. Tscharntke. 2005. Local diversity of arable weeds increases with landscape complexity. Perspectives in Plant Ecology, Evolution and Systematics 7:85–93. Gering, J. C., T. O. Crist, and J. A. Veech. 2003. Additive partitioning of species diversity across multiple spatial scales: implications for regional conservation of biodiversity. Conservation Biology 17:488–499. Grime, J. P. 2001. Plant strategies, vegetation processes, and ecosystem properties. John Wiley, Chichester, UK. Guerrero, I., P. Martı´ nez, M. B. Morales, and J. J. Oñate. 2010. Influence of agricultural factors on weed, carabid and bird richness in a Mediterranean cereal cropping system. Agriculture, Ecosystems and Environment 138:103–108. Hendrickx, F., 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. Holzner, W., and R. Immonen. 1982. Europe: an overview. Pages 203–266 in W. Holzner and M. Numata, editors. Biology and ecology of weeds. Dr. W. Junk, The Hague, Netherlands. Holzschuh, A., I. Steffan-Dewenter, D. Kleijn, and T. Tscharntke. 2007. Diversity of flower-visiting bees in cereal fields: effects of farming system, landscape composition and regional context. Journal of Applied Ecology 44:41–49. Jose´-Marı´ a, L., L. Armengot, J. M. Blanco-Moreno, M. Bassa, and F. X. Sans. 2010. Effects of agricultural intensification on plant diversity in Mediterranean dryland cereal fields. Journal of Applied Ecology 47:832–840. Kleijn, D., et al. 2006. Mixed biodiversity benefits of agrienvironment schemes in five European countries. Ecology Letters 9:243–254.

981

Kleijn, D., et al. 2009. On the relationship between farmland biodiversity and land-use intensity in Europe. Proceedings of the Royal Society B 276:903–909. Kleijn, D., M. Rundlo¨f, J. Scheper, H. G. Smith, and T. Tscharntke. 2011. Does conservation on farmland contribute to halting the biodiversity decline? Trends in Ecology and Evolution 26:474–481. Kleijn, D., and W. J. Sutherland. 2003. How effective are European agri-environment schemes in conserving and promoting biodiversity? Journal of Applied Ecology 40:947–969. Lande, R. 1996. Statistics and partitioning of species diversity, and similarity among multiple communities. Oikos 76:5–13. Lo´pez-Jamar, J., F. Casas, M. Dı´ az, and M. B. Morales. 2011. Local differences in habitat selection by Great Bustards Otis tarda in changing agricultural landscapes: implications for farmland bird conservation. Bird Conservation International 21:328–341. Marshall, E. J. P. 1989. Distribution patterns of plants associated with arable field edges. Journal of Applied Ecology 26:247–257. Marshall, E. J. P. 2009. The impact of landscape structure and sown grass margin strips on weed assemblages in arable crops and their boundaries. Weed Research 49:107–115. Marshall, E. J. P., V. K. Brown, N. D. Boatman, P. J. W. Lutman, G. R. Squire, and L. K. Ward. 2003. The role of weeds in supporting biological diversity within crop fields. Weed Research 43:77–89. Marshall, E. J. P., and A. C. Moonen. 2002. Field margins in northern Europe: their functions and interactions with agriculture. Agriculture, Ecosystems and Environment 89:5–21. Oñate, J. J., J. E. Malo, F. Sua´rez, and B. Peco. 1998. Regional and environmental aspects in the implementation of Spanish agri-environmental schemes. Journal of Environmental Management 52:227–240. Pa¨rtel, M. 2002. Local plant diversity patterns and evolutionary history at the regional scale. Ecology 83:2361–2366. Pinheiro, J. C., and D. M. Bates. 2000. Mixed-effects models in S and S-PLUS. Statistics and computing series. SpringerVerlag, New York, New York, USA. R Development Core Team 2009. R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. http://www.R-project.org Rey-Benayas, J. M., and S. M. Scheiner. 2002. Plant diversity, biogeography and environment in Iberia: patterns and possible causal factors. Journal of Vegetation Science 13:245–258. Roschewitz, I., D. Gabriel, T. Tscharntke, and C. Thies. 2005. The effects of landscape complexity on arable weed species diversity in organic and conventional farming. Journal of Applied Ecology 42:873–882. Stoate, C., N. D. Boatman, R. J. Borralho, C. R. Carvalho, G. R. de Snoo, and P. Eden. 2001. Ecological impacts of arable intensification in Europe. Journal of Environmental Management 63:337–365. Tscharntke, T., A. M. Klein, A. Kruess, I. Steffan-Dewenter, and C. Thies. 2005. Landscape perspectives on agricultural intensification and biodiversity-ecosystem service management. Ecology Letters 8:857–874. Whittingham, M. J. 2007. Will agri-environment schemes deliver substantial biodiversity gain, and if not why not? Journal of Applied Ecology 44:1–5. Wilson, P. J., and N. J. Aebischer. 1995. The distribution of dicotyledonous arable weeds in relation to distance from the field edge. Journal of Applied Ecology 32:295–310.

SUPPLEMENTAL MATERIAL Supplement Plant species found in study areas (Ecological Archives A022-054-S1).