Mar 25, 2014 - 0.173. 0.012. -0.004. 3.50. 0.039. 3.90. 0.032. 0.128. +. 0.170. 0.040 ... 6.60. 0.011. 0.672. +. 0.082. 0.008. 0.020. 6.61. 0.011. 2.555. -0.084.
Appendix S1. – Supplemental Methods
(a) Field margin types: At the local scale, three types of field margins were chosen to represent different levels of structural complexity: (i) grass margin, a perennial grass buffer strip without any nearby tree; (ii) simple hedgerow, a grass buffer strip adjacent to a single storied hedge ( i.e. treeline of full grown trees, formed by a few dominant species, either plane tree Platanus hybrida Brot., white mulberry Morus alba L., or white willow Salix alba L.); (iii) complex hedgerow, a grass buffer strip with a multi-storied hedgerow on the edge (averaged number of tree and shrub species 7.7 ± 2.3; included representative species: elder Sambucus nigra L., common dogwood, blackberry Rubus fruticosus L., Cornus sanguinea L. , field maple Acer campestre L., black locust Robinia pseudoacacia L., common hornbeam Carpinus betulus L., field elm Ulmus minor Mill., or wild cherry Prunus avium L.) (Sitzia et al. 2013). Each field margin was characterized by sampling plant species composition. Sampling was conducted in three plots of 1 × 3 m located along a transect parallel to the field margin. One plot was placed in the middle part of each field margin with the largest side parallel to the field margin. The other two plots were placed at the two margins at least 10 m away from the central plot. For hedgerows, the plots were placed in order to cover both the trees and the grass buffer strip. Vascular plants were recorded to species level (presence/absence data) and once before the first cut of the grass buffer strip at the end of May.
(b) Hedgerow configuration: We calculated the following configuration metrics: (i) number of hedgerow patches within each buffer; (ii) hedgerow edge density within each buffer, calculated as (total edge length) / (buffer area); and (iii) the shape of hedgerows patches, as defined by (patch perimeter) / 4*√(patch area), calculated as the mean of patches within each buffer.
(c) Exclusion experiment: The cages (plastic net with a mesh size of 5 × 5 mm; diameter: 0.3 m; height: 1.20 m) were inserted in a plastic ring (diameter: 0.3 m; height: 0.25 m) dug 10 cm into the
soil. The cages were covered with sticky glue to prevent the access of flying natural enemies. In all cages, a visual inspection was performed to search for and remove by hand any natural enemies before placement of the cages. A pitfall trap filled with a water–detergent mix was added inside the cage to catch the natural enemies that were not detected. Aphids were provided by Katz Biotech AG®. The exclusion experiment started at the beginning of wheat fruit development (the main period of aphid reproduction in the fields). As suggested by Chaplin-Kramer & Kremen (2012), the aphid predation index could underestimate total mortality because it does not account for aphid colonization in the open cages. However, in our case the exclusion experiment started at the beginning of ripening stage, when alate individuals had already disappeared from the crop.
(d) Sampling of ground-dwelling predators: The pitfall traps were filled with 150 ml of 50% ethylene glycol and a plastic roof positioned 10 cm above each trap was used to prevent flooding by rain. The pitfall traps were left out for 10 days concurrently with the exclusion experiment. Invertebrates were preserved in 70% ethanol.
(e) Phytometer experiment: On 25th March 2014, ten seeds of radish were sown into each pot (5-L, 20 cm diameter) filled with clay soil of the Experimental farm of University of Padova. During the leaf development stage, all but two plants out of the ten were removed. Radish plants were cultivated in a greenhouse for 8 weeks (20 May), when they were moved to the field. Before the beginning of the flowering period three pots per field were bagged. Pots were placed into holes in the ground at the field edge, leaving the base of the plants at the level of the ground surface. Plants were periodically watered. Surveys of flower-visiting insects took place between 9:30 and 17:30 under sunny weather conditions with temperatures above 17°C. To avoid any systematic effect of time of the day the sequence of surveys was randomized in the three sampling rounds.
References
Chaplin-Kramer, R. & Kremen, C. (2012) Pest control experiments show benefits of complexity at landscape and local scales. Ecological Applications, 22, 1936–1948. Sitzia, T., Trentanovi, G., Marini, L., Cattaneo, D. & Semenzato, P. (2013) Assessment of hedge stand types as determinants of woody species richness in rural field margins. iForest, 6, 201– 208.
Appendix S2. – Supplemental Tables
Table S1. Mean values ± standard deviation and results of the analysis of variance (ANOVA) testing differences in land cover proportions between the three classes of field margins.
Hedgerow 0.5 km (%) Arable 0.5 km (%) Hedgerow 1 km (%) Arable 1 km (%)
Grass margin 3.0 ± 2.9
Simple hedgerow 3.6 ± 2.7
Complex hedgerow 2.6 ± 1.6
F
P
0.31
0.737
76.5 ± 20.7
70.3 ± 19.4
66.4 ± 18.8
0.60
0.557
2.6 ± 1.8
3.1 ± 2.1
2.6 ± 1.7
0.19
0.827
71.9 ± 18.2
66.6 ± 20.1
64.4 ± 18.8
0.38
0.691
Table S2. Pearson correlations between explanatory variables. Annual plants
Annual plants
Nectar plants 0.502**
Hedgerow (0.5 km)
0.390*
0.086ns
Arable (0.5 km)
0.073ns
0.324 ns
-0.319ns
Hedgerow (1 km)
0.355ns
0.023ns
0.918***
-0.288ns
Arable (1 km)
0.120ns
0.419*
-0.207ns
0.912***
ns
not significant, * P < 0.05, ** P < 0.01, *** P < 0.001
Hedgerow (0.5 km)
Arable (0.5 km)
Hedgerow (1 km)
-0.195 ns
Table S3. Statistical models used in the analyses. Model1
Distribution
Transformation
Fixed effects2
Random effect
Predation index
LMs
Gaussian
log(y+0.01)
FM × AR × H + NP
−
Parasitism rate
GLMMs
Binomial
−
FM × AR × H + AP + NP
Site
Vegetation-dwelling predators
LMMs
Gaussian
log(y+1)
FM × AR × H + AP + NP
Site
Ground-dwelling predators
LMMs
Gaussian
log(y+1)
FM × AR × H + AP + NP
Site
Species richness
GLMs
Poisson
−
FM × AR × H + AN
−
Cover
LMs
Gaussian
log(y+1)
FM × AR × H + AN
−
LMMs
Gaussian
−
FM × AR × H + NP
Site
Response variable Pest control
Weed control
Pollination Visitation rate
Δ Seed set LMMs Gaussian − FM × AR × H + VS + NP Site 1 Model abbreviations: GLMs, generalized linear model; GLMMs, generalized linear mixed models; LMs, linear models; LMMs, linear mixed models. 2 Fixed effect abbreviations: AN, number of annual plant species; AP, aphid abundance; AR, arable land cover; FM, field margin type; H, hedgerow cover; NP, number of nectar plant species; VS, visitation rate.
Table S4. Performance of (a) pest control, (b) weed control, and (c) potential pollination service models using arable land cover and hedgerow cover calculated at the two landscape scales (0.5 or 1 km). Table shows the Akaike Information Criterion (AIC). The most explanatory radius was selected based on minimized AIC values. Landscape scale 0.5 km
1 km
Predation index
-37.22
-36.55
Parasitism rate
270.41
267.97
Vegetation-dwelling
231.68
223.48
Ground-dwelling predators
217.31
218.06
Species richness
137.22
136.35
Cover
185.05
185.75
Visitation rate
374.14
383.09
Δ Seed set
60.86
60.69
(a) Pest control
predators (b) Weed control
(c) Pollination
Table S5. Estimates and model weights of the selected models with ΔAICc < 7.
Aphid abundance
Annual plants richness
Selected models Nectar Visitation plants rate richness
Arable land cover
ΔAICc
Model weight
0.001
0
0.329
0.13
0.308
1.28
0.174
0.001
2.26
0.106
0.002
4.46
0.035
4.94
0.028
-0.008
6.52
0.013
-2.139
0.105
0
0.271
-1.736
0.099
0.71
0.19
1.45
0.131
2.33
0.084
3.08
0.058
3.29
0.052
3.6
0.045
3.89
0.039
4.29
0.032
4.62
0.027
5.22
0.02
-0.007
5.74
0.015
-0.008
6.09
0.013
6.41
0.011
Intercept Predation index
Field margin type
Hedgerow cover
0.828 0.920
Parasitism rate
0.944
-0.008
0.856
-0.005
0.832
+
0.931
+
0.952
+
-1.863
-0.088
0.103
-1.495
-0.082
0.097
-0.006
-0.006
-1.842 -1.399
-0.007
-2.236
+
0.107
-1.779
+
0.101
-1.536
-0.100
-1.126
-0.096
-0.007
-0.006
-1.969
+
-0.084
0.105
-1.555
+
-0.076
0.099
-1.486
+
-1.967
+
Intercept Vegetation-dwelling predators
Ladybirds
Field margin type
Aphid abundance
2.263
0.378
2.403
0.376
2.266
+
0.385
2.397
+
0.382
2.497
0.384
2.322
0.334
2.520
0.332
2.514
0.382
2.283
+
0.371
2.480
+
0.391
Annual plants richness
Selected models Nectar Visitation plants rate richness
Hedgerow cover
Arable land cover
ΔAICc
Model weight
-0.022
0
0.659
-0.023
2.04
0.237
-0.022
4.69
0.063
-0.023
6.9
0.021
-0.025
0.00
0.357
-0.026
0.50
0.279
-0.047
-0.027
2.29
0.114
-0.005
-0.025
2.41
0.107
-0.027
4.23
0.043
-0.025
4.78
0.033
-0.030
6.00
0.018
-0.028
6.31
0.015
0.00
0.226
1.08
0.132
1.62
0.101
2.00
0.083
3.05
0.049
3.10
0.048
3.40
0.041
3.50
0.039
3.90
0.032
4.61
0.023
4.66
0.022
4.80
0.021
-0.033
-0.032 -0.044
-0.043
-0.040
3.304 2.449 Web spiders
+
0.366
0.196
0.198
0.070
0.197
0.451
0.186
0.131
0.188
0.279
0.189
0.239
+
-0.040
0.046 -0.003 0.009 0.041
0.168
0.057
0.195
0.003
0.403
0.173
0.012
0.128
+
0.998 0.647
-0.003
0.170
0.044 -0.004
0.040 -0.005
+
Intercept Web spiders
0.467
Field margin type +
Aphid abundance
Annual plants richness
Selected models Nectar Visitation plants rate richness
0.159
0.524
Arable land cover
ΔAICc
Model weight
-0.003
4.91
0.019
4.98
0.019
5.05
0.018
-0.003
5.36
0.015
-0.006
5.50
0.014
5.52
0.014
-0.004
6.09
0.011
-0.004
6.22
0.010
6.39
0.009
6.73
0.008
6.77
0.008
0.013
0
0.389
0.012
1.52
0.182
2.55
0.109
2.61
0.105
2.75
0.098
4.24
0.047
4.67
0.038
4.98
0.032
0.021
0.00
0.301
0.022
2.14
0.103
0.020
2.14
0.103
0.022
2.29
0.096
0.020
2.42
0.09
0.047
0.482
0.018
0.280
0.181
0.842
0.006
0.035
0.021
0.233
+
0.923
+
0.167
0.001
0.845
0.038
0.549
+
0.322
+
0.037 0.163
0.432 Ground-dwelling predators
Hedgerow cover
0.035 0.013
0.035
1.761 2.024
-0.003
-0.048
2.716 2.957
Carabid beetles
-0.082
1.817
+
0.012
2.589
+
2.013
+
-0.039
2.777
+
-0.069
0.778 0.874
-0.013
0.921
-0.026
0.752 0.902
0.010 +
0.010
Intercept Carabid beetles
0.707
Field margin type +
0.820
Aphid abundance
Annual plants richness
Selected models Nectar Visitation plants rate richness
0.949 0.019 +
0.944
+
0.021
4.15
0.038
0.022
4.40
0.033
-0.019
0.021
4.43
0.033
-0.028
0.021
4.47
0.032
0.019
4.64
0.03
0.020
4.84
0.027
5.84
0.016
0.020
6.60
0.011
0.020
6.61
0.011
6.65
0.011
0.021
6.77
0.01
-0.012
0.00
0.212
1.25
0.113
-0.012
1.90
0.082
-0.011
2.08
0.075
-0.012
2.23
0.07
2.71
0.055
2.81
0.052
3.15
0.044
3.24
0.042
-0.010
3.51
0.037
-0.011
3.99
0.029
4.27
0.025
4.41
0.023
-0.008
+
0.093
0.672
+
0.082
-0.015 0.008
2.555
Rove beetles
Model weight
0.016
0.777
0.896
ΔAICc
-0.018 -0.009
0.796
Arable land cover
0.089 0.036
0.885
Hedgerow cover
-0.084 0.037
-0.014
-0.019
1.867
1.990 1.747
-0.017 0.042
1.795 0.901
0.018 0.081
1.252
-0.024
0.921
0.042
1.189
0.136
-0.041
1.831
0.087
-0.030
1.875
-0.023
0.039
1.146
-0.033
0.069
1.704
0.039
0.013
-0.011
Intercept Rove beetles
1.872
Field margin type +
Aphid abundance
Annual plants richness
Selected models Nectar Visitation plants rate richness
Hedgerow cover
Arable land cover
ΔAICc
Model weight
-0.013
4.50
0.022
0.031
4.83
0.019
4.91
0.018
5.67
0.012
5.76
0.012
-0.012
6.45
0.008
-0.012
6.69
0.007
-0.012
6.83
0.007
0.00
0.219
1.35
0.112
1.81
0.089
1.89
0.085
-0.018
1.99
0.081
-0.034
2.90
0.051
-0.003
3.08
0.047
-0.004
3.56
0.037
3.60
0.036
3.76
0.034
-0.002
3.88
0.032
-0.005
3.91
0.031
4.41
0.024
5.17
0.017
-0.003
5.45
0.014
-0.003
5.51
0.014
0.832
0.073
1.098
0.128
-0.048
0.062
1.717
0.086
-0.035
0.039
1.059
+
1.701
+
1.972
+
1.787
+
0.068 -0.016 0.022
-0.009
Cursorial spiders 0.646
0.018
1.003
-0.003
0.748
0.035
0.861 0.693
0.024
0.858
0.019
1.162
-0.029
0.640
0.013
0.806
0.045
0.928
0.029
1.070 0.838
0.016 -0.025
0.029
-0.053
+
0.687
0.020
0.021
0.857
0.001
0.019
1.088
0.040
-0.036
-0.034
Intercept Cursorial spiders
0.649
Field margin type +
1.066
Weed species richness
Aphid abundance
Annual plants richness
0.003
1.004
+
0.895
+
0.780
+
Selected models Nectar Visitation plants rate richness 0.021 0.028
Hedgerow cover
-0.053
ΔAICc
Model weight
5.88
0.012
-0.005
6.35
0.009
-0.003
6.43
0.009
6.47
0.009
6.48
0.009
-0.010
0
0.382
-0.012
1.96
0.143
-0.010
2.2
0.127
2.56
0.106
3.89
0.055
-0.011
4.12
0.049
-0.011
4.56
0.039
4.91
0.033
6.45
0.015
6.73
0.013
6.79
0.013
6.86
0.012
0
0.377
1.94
0.143
2.32
0.118
2.52
0.107
3.53
0.065
0.091
4.07
0.049
0.082
4.48
0.04
-0.021 0.036
2.480 2.462
Arable land cover
0.024
2.369
0.029
1.812 1.680 2.467
Weed cover
0.046 +
2.372
0.022
1.833
-0.004
1.705
-0.005
2.345
+
1.792
+
2.464
+
0.023
0.046 0.035
0.020
-0.010
-0.012
8.564 3.815
0.066
7.490
0.342
9.290 6.263
-0.146 +
0.227 4.278
0.584 -0.318
Intercept Weed cover
Field margin type
8.279 2.158
+
7.275
+
5.835
+
0.180
Visitation rate
2.356
Aphid abundance
Annual plants richness -0.169
Selected models Nectar Visitation plants rate richness
ΔAICc
Model weight
5.08
0.03
6.1
0.018
6.45
0.015
6.62
0.014
6.63
0.014
0.935
0
0.293
0.891
0.06
0.284
-0.057
0.967
2.19
0.098
0.028
0.878
2.31
0.092
0.062 -0.261 0.162 -0.436
0.695
+
+
2.636 2.730 2.037
+
2.247
+
2.621 Δ Seed set
0.002
2.38
0.089
0.948
0.004
2.44
0.086
-0.070
1.000
0.009
4.6
0.029
0.028
0.879
0.000
4.71
0.028
0.115
0
0.245
0.137
0.76
0.168
0.102
1.99
0.09
2.2
0.082
2.82
0.06
2.99
0.055
3.34
0.046
3.46
0.043
4.02
0.033
4.67
0.024
0.846
-0.033 0.024
0.347
0.128
0.643
0.117
0.895
0.707
0.518
Arable land cover
0.358
2.850 2.795
Hedgerow cover
0.027
-0.033
0.004
0.123
1.044 0.539
-0.031
0.640 0.910 0.285
0.147
0.004
0.043 +
0.125 0.020
0.114
0.003
Intercept Δ Seed set
Field margin type
1.164 1.032
Aphid abundance
Annual plants richness
Selected models Nectar Visitation plants rate richness -0.020
+
0.741
-0.031 0.047
Hedgerow cover
ΔAICc
Model weight
4.96
0.021
5.36
0.017
5.4
0.016
-0.001
5.53
0.015
0.149
0.008
5.56
0.015
0.131
0.002
5.87
0.013
-0.002
6.08
0.012
0.111
6.61
0.009
0.068
0
0.376
2.14
0.129
2.27
0.121
2.28
0.12
3.7
0.059
4.32
0.043
4.61
0.038
-0.001
4.64
0.037
-0.003
5.87
0.02
6.37
0.016
0
0.379
2.06
0.135
2.24
0.124
2.32
0.119
4.39
0.042
0.146
-0.022
1.130 0.351
0.023
0.798
0.045
0.723 Hoverflies
+
0.471
+
0.023
0.801
Arable land cover
-0.032
0.992 0.871
0.065
0.827
-0.003
0.069
1.214
-0.003
0.948 0.801
Wild bees
-0.001
0.005 +
0.066
0.882
-0.003
1.167
0.007
0.976
+
0.078
+
0.306
+
-0.092
+
0.148
+
-0.009
0.248
+
0.006
0.067
0.086
0.093 0.091
0.002
0.386
Field margin type +
-0.050
+
Intercept Wild bees
Aphid abundance
Annual plants richness
Selected models Nectar Visitation plants rate richness -0.014
0.306
Hedgerow cover
Arable land cover
ΔAICc
Model weight
-0.001
4.4
0.042
0.003
4.46
0.041
5.21
0.028
5.25
0.027
6.7
0.013
6.78
0.013
0.085
0
0.268
0.081
1.17
0.15
2.27
0.086
2.36
0.082
2.76
0.067
2.85
0.065
3.06
0.058
3.37
0.05
4.24
0.032
4.46
0.029
-0.003
4.71
0.025
-0.001
4.78
0.024
5.16
0.02
-0.003
5.2
0.02
0.103 0.074
0.515 0.311
Butterflies
0.023
0.334
+
0.321
+
0.008
0.163 0.423
+
0.361
+
-0.001
0.082 -0.005
-0.001
0.088
0.376 0.543
+
0.336
0.077
0.121
0.006
-0.002
0.078
0.608
-0.003
0.263
0.013
0.713
+
0.436
+
-0.004
0.499
+
0.005
0.085
0.303
0.008
0.511
0.016
-0.004
5.7
0.016
0.008
-0.003
6.98
0.008
0.665
+
0.072
Intercept Other groups
Field margin type
Aphid abundance
Annual plants richness
Selected models Nectar Visitation plants rate richness
Hedgerow cover
Arable land cover
-0.212 -0.471
0.092
-0.034
-0.020
-0.395
0.003
-0.213
-0.036
-0.978 -0.809 -0.231
-0.043
0.110
+
-0.401
+
-0.146
+
-0.919
+
-0.217
+
0.267
0.72
0.186
1.95
0.1
2.1
0.093
2.28
0.085
2.45
0.078
0.135
0.008
3.75
0.041
3.76
0.041
4.09
0.035
4.94
0.023
6
0.013
6.04
0.013
6.84
0.009
6.92
0.008
0.003 0.082 0.003
-0.009 0.101 -0.031
0
0.007
-0.022
-0.441
Model weight
0.109
+
-0.241
ΔAICc
0.103
0.007
Table S6. Sum of model weights (Σwi) for each variable estimated by the multi-model inference procedure for vegetation-dwelling and ground-dwelling predators. In bold are reported the Akaike weights of the most important explanatory variables (sum of model weights > 0.6 or unconditional CIs that did not include 0). The direction of the relationship is indicated by (+) or (-) for the most important continuous variables. Field margin
Aphid abundance
Ladybirds
0.26
0.96 (+)
0.51
1.00 (−)
0.26
Web spiders
0.33
0.80 (+)
0.27
0.31
0.34
Carabid beetles
0.24
0.25
0.25
0.93 (+)
0.26
Rove beetles
0.09
0.31
0.32
0.59
0.26
Cursorial spiders
0.09
0.26
0.35
0.30
0.30
Nectar plant Arable land richness cover
Hedgerow cover
Vegetation-dwelling predators
Ground-dwelling predators
Table S7. Sum of model weights (Σwi) for each variable estimated by the multi-model inference procedure for pollinator groups (abundance of flower-visiting insects). In bold are reported the Akaike weights of the most important explanatory variables (sum of model weights > 0.6 or unconditional CIs that did not include 0). The direction of the relationship is indicated by (+) or (-) for the most important continuous variables. Field margin
Nectar plant richness
Arable land cover
Hedgerow cover
Hoverflies
0.09
0.24
0.26
0.72 (+)
Wild bees
0.90
0.25
0.25
0.71 (+)
Butterflies
0.58
0.25
0.27
0.74 (+)
0.28
0.43
Pollinator groups
Other groups1 0.11 0.29 1 Including other Diptera and other species (mainly coleopterans).
Appendix S3. – Supplemental Figures
Figure S1. Scatterplot showing the distribution of the three field margin types along the two landscape gradients: (i) proportion of arable land and (ii) proportion of hedgerows in a 1 km buffer.
Figure S2. Mean (± SE) number of aphids, Sitobion avenae, per pot in the cage and the open treatments at the beginning (day 0) and the end (day 5) of the experiment. No significant differences were found between the two treatments at day 0 (Paired t-test: t = 1.66, P = 0.108), while significant differences were found between the cage and the open treatment at day 5 (Paired t-test: t = 11.93, P < 0.001).
Figure S3. Total abundance (number of individuals) of flower-visiting insects surveyed during the three visits.
Figure S4. Model-averaged effect sizes and 95% confidence intervals for the landscape parameters included in strongly supported models (ΔAICc ≤ 7) explaining the abundance of flower-visiting insects for (a) hoverflies, (b) wild bees, (c) butterflies, and (d) other groups (other Diptera and other species - mainly coleopterans). Open and closed symbols reflect model estimates ± averaged confidence intervals.
Figure S5. Mean ± SE of (a) wild bee and (b) butterfly abundance derived from visitation rate in relation to margin type (grass margin, simple hedgerow, and complex hedgerow).