1 A. Salisbury et al. 2017. Enhancing gardens as habitats for invertebrates Online Resource 2: Further details of Results Enhancing gardens as habitats for plant-associated invertebrates: should we plant native or exotic species? Andrew Salisbury1*, Sarah Al-Beidh1, James Armitage1, Stephanie Bird1/2, Helen Bostock 1, Anna Platoni1, Mark Tatchell3, Ken Thompson4, Joe Perry5 1 Royal 2 Horticultural Society, RHS Garden Wisley, Woking, Surrey, GU23 6QB, UK Department of Life Sciences, University of Roehampton, Whitelands College, Holybourne Avenue, London, SW15 4JD 3Laurels 4 Farm, Oborne, Sherborne, Dorset, DT9 4LA, UK Department of Animal and Plant Sciences, University of Sheffield, Western Bank, Sheffield, S10 2TN, UK 5 Oaklands Barn, Lug’s Lane, Broome, Norfolk, NR35 2HT, UK *E-mail: [email protected] Contents Table S2.1: Fitted regressions for total abundance Herbivores for native (N), 4 near-native (Z) and exotic (E) treatments Table S2.2: Fitted regressions for abundance of Herbivores (sucking) for native 6 (N), near-native (Z) and exotic (E) treatments. Table S2.3: Fitted regressions for abundance of Herbivores (chewing) for 8 native (N), near-native (Z) and exotic (E) treatments. Table S2.4: Fitted regressions for abundance of Herbivore (generalist) for 10 native (N), near-native (Z) and exotic (E) treatments. Table S2.5: Fitted regressions for abundance of Herbiovore (specalist) for 12 native (N), near-native (Z) and exotic (E) treatments. Table S2.6: Fitted regressions for abundance of Predators for native (N), near- 14 2 A. Salisbury et al. 2017. Enhancing gardens as habitats for invertebrates native (Z) and exotic (E) treatments. Table S2.7: Fitted regressions for abundance of Predators (excluding 16 Araenidae & Parasitica) for native (N), near-native (Z) and exotic (E) treatments. Table S2.8: Fitted regressions for abundance Parasitica for native (N), near- 18 native (Z) and exotic (E) treatments. Table S2.9: Fitted regression of abundance Araneae (web-spinners) for native 20 (N), near-native (Z) and exotic (E) treatments. Table S2.10: Fitted regressions for abundance of Araneae (hunters) for native 22 (N), near-native (Z) and exotic (E) treatments. Table S2.11: Fitted regressions for abundance Detritivoure/fungivoures for 24 native (N), near-native (Z) and exotic (E) treatments. Table S2.12: Fitted regressions for abundance of Omnivoures for native (N), 26 near-native (Z) and exotic (E) treatments. Fig. S2.1: Dependence of abundance of Herbivores on plant canopy. 5 Fig. S2.2: Dependence of abundance of Herbivores (sucking) on plant 7 canopy. Fig. S2.3: Dependence of abundance of plant inhabiting invertebrate 9 Herbivores (chewing) on plant canopy. Fig. S2.4: Dependence of abundance of Herbivore (generalist) on plant 11 canopy. Fig. S2.5: Dependence of Herbivore (specialist) on plant canopy. 13 Fig. S2.6: Dependence of abundance of Predators on plant canopy. 15 Fig. S2.7: Dependence of abundance of Predators (excluding Araneidae & 17 Parasitica) on plant canopy. Fig. S2.8: Dependence of abundance of Parasitica on plant canopy. 19 Fig. S2.9: Lack of dependence of abundance of Araneae (web-spinners) on 21 canopy cover. Fig. S2.10: Dependence of Araneae (hunters) on canopy cover. 23 Fig. S2.11: Dependence of abundance of Detritivores on plant canopy. 25 3 A. Salisbury et al. 2017. Enhancing gardens as habitats for invertebrates Fig. S2.12: Dependence of abundance of Omnivores on plant canopy. 27 4 A. Salisbury et al. 2017. Enhancing gardens as habitats for invertebrates Appendix S1: Further details of Results Table S2.1 Fitted regressions for total abundance of Herbivores for native (N), near-native (Z) and exotic (E) treatments. The estimated intercepts and slope are fitted values from the best-fitting model, three parallel lines. F intercept is the F-statistic for the test of differences between the intercepts of the linear relationships for the three treatments, with probability level given by P intercept. All F-statistics for intercepts have 2,m degrees of freedom where m>100. No corresponding values of F-statistics are given for slopes because these were always non-significant. Estimated Treatment F P Estimated intercept intercept slope 0.058 1.66 0.19 0.20 0.12 0.084 2.38 0.095 0.52 0.12 0.083 1.97 0.14 0.36 0.11 0.12 1.89 0.15 0.63 0.14 0.039 4.86 0.008 0.51 0.54 SE intercept SE 2010 N 0.37 Z 0.27 E 0.28 N 0.29 Z 0.28 E 0.16 N 0.23 Z 0.28 E 0.17 N 0.31 Z 0.22 E 0.16 2011 2012 2013 All years combined N 0.25 Z 0.22 E 0.16 A. Salisbury et al. 2017. Enhancing gardens as habitats for invertebrates 5 Fig. S2.1. Dependence of abundance of Herbivores on plant canopy. Observed data and fitted regressions for Herbivores (n = 3 273), over all years 2010–2013 combined. Native (N) green, downward triangles; near-native (Z) blue, squares; exotic (E) red, upward triangles. Estimated intercepts and slope are shown in Table S2.1. The fitted regressions (intercepts) differ (F 3,743 = 4.86, P =0.008). 6 A. Salisbury et al. 2017. Enhancing gardens as habitats for invertebrates Table S2.2. Fitted regressions for abundance of Herbivores (sucking) for native (N), near-native (Z) and exotic (E) treatments. The estimated intercepts and slope are fitted values from the best-fitting model, three parallel lines. F intercept is the F-statistic for the test of differences between the intercepts of the linear relationships for the three treatments, with probability level given by P intercept. All F-statistics for intercepts have 2,m degrees of freedom where m>100. No corresponding values of F-statistics are given for slopes because these were always non-significant. F P interce intercept Estimated Treatment Estimated SE intercept SE slope pt 2010 N 0.29 Z 0.16 E 0.23 N 0.26 Z 0.22 E 0.14 N 0.27 Z 0.29 E 0.19 N 0.24 Z 0.13 E 0.093 0.059 2.28 0.11 0.19 0.12 1.82 0.16 0.51 0.13 0.084 1.90 0.15 0.23 0.11 0.12 1.76 0.18 0.64 0.15 0.040 3.99 0.019 0.50 0.056 2011 0.085 2012 2013 All years combined N 0.20 Z 0.15 E 0.11 A. Salisbury et al. 2017. Enhancing gardens as habitats for invertebrates 7 Fig. S2.2. Dependence of abundance of Herbivores (sucking) on plant canopy. Observed data and fitted regressions for Herbivores (sucking) (n = 2809), over all years 2010–2013 combined. Native (N) green, downward triangles; near-native (Z) blue, squares; exotic (E) red, upward triangles. Estimated intercepts and slope are shown in Table S2.2. The fitted regressions (intercepts) differ (F3,743 = 3.99, P =0.019). 8 A. Salisbury et al. 2017. Enhancing gardens as habitats for invertebrates Table S2.3. Fitted regressions for abundance of Herbivores (chewing) for native (N), near-native (Z) and exotic (E) treatments. The estimated intercepts and slope are fitted values from the best-fitting model, three parallel lines. F intercept is the F-statistic for the test of differences between the intercepts of the linear relationships for the three treatments, with probability level given by P intercept. All F-statistics for intercepts have 2,m degrees of freedom where m>100. No corresponding values of F-statistics are given for slopes because these were always non-significant. Estimated Treatment F P Estimated intercept intercept slope 2.77 0.065 -0.015 0.067 0.042 1.97 0.14 0.15 0.063 0.049 0.33 0.72 0.29 0.066 0.072 2.13 0.12 0.16 0.086 0.22 3.80 0.023 0.14 0.030 SE intercept SE 2010 N 0.17 Z 0.17 E 0.088 N 0.012 Z 0.050 E -0.012 N -0.099 Z -0.074 E -0.079 N 0.094 Z 0.086 E 0.0044 0.034 2011 2012 2013 All years combined N 0.054 Z 0.065 E 0.019 A. Salisbury et al. 2017. Enhancing gardens as habitats for invertebrates 9 Fig. S2.3. Dependence of abundance of plant inhabiting invertebrate Herbivores (chewing) on plant canopy. Observed data and fitted regressions for plant inhabiting invertebrate Herbivores (chewing) (n = 463), over all years 2010–2013 combined. Native (N) green, downward triangles; near-native (Z) blue, squares; exotic (E) red, upward triangles. Estimated intercepts and slope are shown in Table S2.3. The fitted regressions (intercepts) differ (F3,743 = 3.80, P =0.023). 10 A. Salisbury et al. 2017. Enhancing gardens as habitats for invertebrates Table S2.4. Fitted regressions for abundance of Herbivore (generalist) for native (N), near-native (Z) and exotic (E) treatments. The estimated intercepts and slope are fitted values from the best-fitting model, three parallel lines. F intercept is the F-statistic for the test of differences between the intercepts of the linear relationships for the three treatments, with probability level given by P intercept. All F-statistics for intercepts have 2,m degrees of freedom where m>100. No corresponding values of F-statistics are given for slopes because these were always non-significant. Estimated Treatment F P Estimated intercept intercept slope 0.035 2.40 0.094 0.14 0.069 0.040 0.39 0.68 0.18 0.060 0.038 0.54 0.59 0.059 0.051 0.080 2.42 0.093 0.39 0.096 0.022 4.09 0.017 0.14 0.030 SE intercept SE 2010 N 0.11 Z 0.050 E 0.039 N -0.010 Z -0.029 E -0.037 N 0.037 Z 0.012 E 0.020 N -0.079 Z -0.062 E -0.17 2011 2012 2013 All years combined N 0.046 Z 0.18 E -0.005 A. Salisbury et al. 2017. Enhancing gardens as habitats for invertebrates 11 Fig. S2.4. Dependence of abundance of Herbivore (generalist) on plant canopy. Observed data and fitted regressions for plant inhabiting invertebrate Herbivore (generalist) (n = 387), over all years 2010–2013 combined. Native (N) green, downward triangles; near-native (Z) blue, squares; exotic (E) red, upward triangles. Estimated intercepts and slope are shown in Table. S2.4. The fitted regressions (intercepts) differ (F3,743 = 4.09, P =0.017). 12 A. Salisbury et al. 2017. Enhancing gardens as habitats for invertebrates Table S2.5. Fitted regressions for abundance of Herbiovore (specalist) for native (N), near-native (Z) and exotic (E) treatments. Differences between intercepts were tested by partial F-tests (F int. with probablility P int.); similarly for slopes (F slopes with probablility P slopes). All F-statistics have 2,m degrees of freedom where m>100. For almost all tests, values of both F statistics are non-significant, indicating that the best-fitting model is a single line with no significant differences between the treatments. F P F slopes P slopes Estimated Treatment Estimate SE intercept SE d slope int. int. 0.58 0.56 1.72 0.18 0.093 0.029 0.006 0.061 0.26 0.77 1.35 0.26 -0.052 0.035 0.20 0.051 1.47 0.23 0.66 0.52 -0.092 0.41 0.26 0.057 1.10 0.34 0.28 0.76 -0.0066 0.062 0.22 0.079 0.15 1.77 0.17 0.002 0.018 0.16 0.027 2010 N Z E 2011 N Z E 2012 N Z E 2013 N Z E All years combined N Z E 1.90 A. Salisbury et al. 2017. Enhancing gardens as habitats for invertebrates 13 Fig. S2.5. Dependence of Herbivore (specialist) on plant canopy. Observed data and fitted regressions for plant inhabiting invertebrate Herbivore (specialist) abundance (n = 338), over all years 2010–2013 combined. Native (N) green, downward triangles; near-native (Z) blue, squares; exotic (E) red, upward triangles. Estimated intercept and slope are shown in Table S2.5. The fitted regression is significant (F1,746 = 33.04, P < 0.001) and is the same for all three treatments. 14 A. Salisbury et al. 2017. Enhancing gardens as habitats for invertebrates Table S2.6. Fitted regressions for abundance of Predators for native (N), near-native (Z) and exotic (E) treatments. The estimated intercepts and slope are fitted values from the best-fitting model, three parallel lines. F intercept is the F-statistic for the test of differences between the intercepts of the linear relationships for the three treatments, with probability level given by P intercept. All F-statistics for intercepts have 2,m degrees of freedom where m>100. No corresponding values of F-statistics are given for slopes because these were always non-significant. Estimated Treatment F intercept P Estimated intercept slope SE SE intercept 2010 N 0.47 Z 0.35 E 0.33 N 0.30 Z 0.28 E 0.29 N 0.61 Z 0.48 E 0.50 N 0.79 Z 0.70 E 0.59 0.047 5.19 0.007 0.37 0.093 0.85 0.07 0.93 0.65 0.085 0.068 4.79 0.009 0.012 0.091 0.071 9.17 <0.001 -0.038 0.86 0.029 11.66 <0.001 0.27 0.041 2011 2012 2013 All years combined N 0.51 Z 0.43 E 0.39 A. Salisbury et al. 2017. Enhancing gardens as habitats for invertebrates 15 Fig. S2.6. Dependence of abundance of Predators on plant canopy. Observed data and fitted regressions for plant inhabiting invertebrate Predators on plant canopy (n = 3143 ), over all years 2010–2013 combined. Native (N) green, downward triangles; near-native (Z) blue, squares; exotic (E) red, upward triangles. Estimated intercepts and slope are shown in Table S2.6. The fitted regressions (intercepts) differ (F3,743 = 11.66, P <0.001). 16 A. Salisbury et al. 2017. Enhancing gardens as habitats for invertebrates Table S2.7. Fitted regressions for abundance of Predators (excluding Araenidae & Parasitica) for native (N), near-native (Z) and exotic (E) treatments. The estimated intercepts and slope are fitted values from the best-fitting model, three parallel lines. F intercept is the F-statistic for the test of differences between the intercepts of the linear relationships for the three treatments, with probability level given by P intercept. All F-statistics for intercepts have 2,m degrees of freedom where m>100. No corresponding values of F-statistics are given for slopes because these were always nonsignificant. Estimated Treatment F P Estimated intercept intercept slope 0.035 2.70 0.070 0.042 0.069 0.044 0.84 0.43 0.32 0.065 0.037 3.49 0.032 0.14 0.037 0.071 5.05 0.008 0.035 0.086 0.021 6.18 0.002 0.12 0.030 SE intercept SE 2010 N 0.14 Z 0.11 E 0.0059 N -0.12 Z -0.15 E -0.14 N 0.12 Z 0.0011 E 0.023 N 0.20 Z 0.20 E 0.068 2011 2012 2013 All years combined N 0.071 Z 0.050 E 0.010 A. Salisbury et al. 2017. Enhancing gardens as habitats for invertebrates 17 Fig. S2.7. Dependence of abundance of Predators (excluding Araneidae & Parasitica) on plant canopy. Observed data and fitted regressions for plant inhabiting invertebrate predators (excluding Araneidae & Parasitica) (n = 395), over all years 2010–2013 combined. Native (N) green, downward triangles; near-native (Z) blue, squares; exotic (E) red, upward triangles. Estimated intercepts and slope are shown in Table S2.7. The fitted regressions (intercepts) differ (F3,743 = 6.18, P = 0.002). 18 A. Salisbury et al. 2017. Enhancing gardens as habitats for invertebrates Table S2.8. Fitted regressions for abundance Parasitica for native (N), near-native (Z) and exotic (E) treatments. The estimated intercepts and slope are fitted values from the best-fitting model, three parallel lines. F intercept is the F-statistic for the test of differences between the intercepts of the linear relationships for the three treatments, with probability level given by P intercept. All F-statistics for intercepts have 2,m degrees of freedom where m>100. No corresponding values of F-statistics are given for slopes because these were always non-significant. Estimated Treatment F P Estimated intercept intercept slope 12.06 <0.001 0.14 0.092 1.02 0.36 0.77 0.099 0.061 3.68 0.027 0.29 0.74 4.83 0.009 0.23 0.089 0.029 16.47 <0.001 0.33 0.041 SE intercept SE 2010 N 0.31 Z 0.10 E 0.13 N -0.10 Z -0.58 E -0.13 N 0.12 Z 0.025 E 0.018 N 0.21 Z 0.10 E 0.06 0.47 2011 0.067 2012 0.082 2013 All years combined N 0.17 Z 0.052 E 0.040 A. Salisbury et al. 2017. Enhancing gardens as habitats for invertebrates 19 Fig. S2.8. Dependence of abundance of Parasitica on plant canopy. Observed data and fitted regressions for plant inhabiting Parasitica (n = 1137), over all years 2010–2013 combined. Native (N) green, downward triangles; near-native (Z) blue, squares; exotic (E) red, upward triangles. Estimated intercepts and slope are shown in Table S2.8. The fitted regressions (intercepts) differ (F3,743 = 16.47, P < 0.001). 20 A. Salisbury et al. 2017. Enhancing gardens as habitats for invertebrates Table S2.9. Fitted regression of abundance Araneae (web-spinners) for native (N), near-native (Z) and exotic (E) treatments. Differences between intercepts were tested by partial F-tests (F int. with probablility P int.); similarly for slopes (F slopes with probablility P slopes). All F-statistics have 2,m degrees of freedom where m>100. Values of both F statistics are non-significant, indicating that the best-fitting model is a single line with no significant differences between the treatments. The regression is not significant (F1,745 = 0.94, P = 0.332). Similar results obtained for individual years, not shown. F P F P Estimated Treatment Estimated SE int. int. slopes slopes intercept 0.76 0.14 0.87 0.27 SE slope All years combined N Z E 0.27 0.025 0.037 0.038 A. Salisbury et al. 2017. Enhancing gardens as habitats for invertebrates 21 Fig.S2.9. Lack of dependence of abundance of Araneae (web-spinners) on canopy cover. Observed data and fitted regressions for plant inhabiting Araneae (web spinners) abundance (n = 1076), over all years 2010–2013 combined. Native (N) green, downward triangles; near-native (Z) blue, squares; exotic (E) red, upward triangles. The regression is not significant (F1, 746 = 0.94, P = 0.332). Estimates of intercept and slope are given in Table S2.9. 22 A. Salisbury et al. 2017. Enhancing gardens as habitats for invertebrates Table S2.10. Fitted regressions for abundance of Araneae (hunters) for native (N), near-native (Z) and exotic (E) treatments. Differences between intercepts were tested by partial F-tests (F int. with probablility P int.); similarly for slopes (F slopes with probablility P slopes). All F-statistics have 2,m degrees of freedom where m>100. For almost all tests, values of both F statistics are non-significant, indicating that the best-fitting model is a single line with no significant differences between the treatments. F P F slopes P slopes Estimated Treatment Estimate SE intercept SE d slope int. int. 0.19 0.83 0.03 0.97 0.047 0.30 0.31 0.077 1.30 0.27 1.17 0.31 0.044 0.051 0.24 0.073 1.35 0.26 1.42 2.63 0.074 0.10 -0.006 0.059 0.25 0.78 0.30 0.74 0.21 0.06 -0.026 0.077 0.96 0.35 0.71 0.11 0.022 0.087 0.032 2010 N Z E 2011 N Z E 2012 N Z E 2013 N Z E All years combined N Z E 0.04 A. Salisbury et al. 2017. Enhancing gardens as habitats for invertebrates 23 Fig. S2.10. Dependence of Araneae (hunters) on canopy cover. Observed data and fitted regressions for plant inhabiting Araneae (hunters) abundance (n = 1 611), over all years 2010–2013 combined. Native (N) green, downward triangles; near-native (Z) blue, squares; exotic (E) red, upward triangles. Estimated intercepts and slope are shown in Table S2.10. The fitted regression is significant (F1, 746 = 7.31, P = 0.007) and is the same for all three treatments. 24 A. Salisbury et al. 2017. Enhancing gardens as habitats for invertebrates Table S2.11. Fitted regressions for abundance Detritivoure/fungivoures for native (N), near-native (Z) and exotic (E) treatments. The estimated intercepts and slope are fitted values from the best-fitting model, three parallel lines. F intercept is the F-statistic for the test of differences between the intercepts of the linear relationships for the three treatments, with probability level given by P intercept. All F-statistics for intercepts have 2,m degrees of freedom where m>100. No corresponding values of F-statistics are given for slopes because these were always non-significant. Estimated Treatment F P Estimated intercept intercept slope 3.45 0.034 -0.042 0.16 0.090 7.51 <0.001 0.45 0.13 0.13 4.64 0.011 -1.05 0.17 1.44 4.73 0.010 -0.20 0.17 0.047 9.78 <0.001 0.26 0.077 SE intercept SE 2010 N 0.72 Z 0.52 E 0.56 N 0.72 Z 0.47 E 0.62 N 1.77 Z 1.55 E 1.75 N 1.45 Z 1.16 E 1.36 0.082 2011 2012 2013 All years combined N 0.82 Z 0.63 E 0.76 A. Salisbury et al. 2017. Enhancing gardens as habitats for invertebrates 25 Fig. S2.11. Dependence of abundance of Detritivores on plant canopy. Observed data and fitted regressions for plant inhabiting Detritivore/fungivore invertebrates (n = 10 982), over all years 2010– 2013 combined. Native (N) green, downward triangles; near-native (Z) blue, squares; exotic (E) red, upward triangles. Estimated intercepts and slope are shown in Table S2.10. The fitted regressions (intercepts) differ (F 3,743 = 9.78, P < 0.001). 26 A. Salisbury et al. 2017. Enhancing gardens as habitats for invertebrates Table S2.12. Fitted regressions for abundance of Omnivoures for native (N), near-native (Z) and exotic (E) treatments. The estimated intercepts and slope are fitted values from the best-fitting model, three parallel lines. F intercept is the F-statistic for the test of differences between the intercepts of the linear relationships for the three treatments, with probability level given by P intercept. All F-statistics for intercepts have 2,m degrees of freedom where m>100. No corresponding values of F-statistics are given for slopes because these were always non-significant. Estimated Treatment F P Estimated intercept intercept slope 0.06 0.94 0.47 0.092 9.43 <0.001 0.56 0.085 3.31 0.038 0.24 0.058 0.41 0.67 0.080 0.069 0.026 5.39 0.005 0.23 0.036 SE intercept SE 2010 N -0.032 Z -0.024 E -0.016 N -0.25 Z -0.065 E -0.17 N -0.020 Z -0.0098 E -0.087 N 0.045 Z 0.069 E 0.079 0.044 2011 0.058 2012 0.051 0.068 2013 All years combined N -0.0038 Z 0.054 E -0.006 27 A. Salisbury et al. 2017. Enhancing gardens as habitats for invertebrates Fig. S2.12. Dependence of abundance of Omnivores on plant canopy. Observed data and fitted regressions for everything (n = 572), over all years 2010–2013 combined. Native (N) green, downward triangles; near-native (Z) blue, squares; exotic (E) red, upward triangles. Estimated intercepts and slope are shown in Table S2.12. The fitted regressions (intercepts) differ (F P = 0.005). 3,743 = 5.39,
© Copyright 2026 Paperzz