Report Specialization of Mutualistic
1. Supplemental Data
Figure S1.
Dissertation J Fründ 112
Figure S1, Related to Figure 1. Latitudinal Specialization Trends in Standardized and Unstandardized Network Metrics
(A) Network specialization H2', i.e., standardized Shannon entropy, (B) connectance, i.e., the realized proportion of possible links, (C) unweighted generality, i.e., the average number of links (the number of observed resource plant species) per consumer species, (D) weighted generality, i.e., the average effective number of links per consumer species (accounting for interaction strength), (E) plant specialization di', and (F) animal specialization dj'. For (E) and (F) partial residuals are shown because regression models were adjusted for the effects of mean web asymmetry on plant and animal specialization in each region; web asymmetry was given as the difference between the effective number of plant and animal species standardized by the sum of the effective number of plant and animal species. Results for (E) and (F) were qualitatively identical for weighted and unweighted means of d', shown are weighted means across species.
Symbol size corresponds to weights by sampling intensity in each region. We focus on a null-model adjusted version of (A) in the main text because it was the only metric that was independent of sampling effort and network size (Table S3). All network metrics showed the same trend: Tropical communities were more generalized than temperate communities.
Dissertation J Fründ 113
Figure S2, Related to Figure 2. Relationships between Cumulative Annual Temperature (Growing Degree Days) and Other Climatic Variables
(A) Climate-change velocity (log10-scale), (B) annual precipitation, (C) potential evapotranspiration (PET) and (D) actual evapotranspiration (AET). Red triangles indicate regions with pollination networks, blue triangles regions with seed dispersal networks. Filled triangles indicate tropical regions, open triangles indicate non-tropical regions. Cumulative annual temperature is closely related to gradients in annual precipitation, AET and PET, probably because climates in most study regions were not limited by water availability. Values for potential (PET) and actual evapotranspiration (AET) were taken from a global aridity database (http://www.cgiar-csi.org/data/item/51-global-aridity-and-pet-database).
Dissertation J Fründ 114
Figure S3, Related to Figure 3. Relationship between Plant Diversity and Latitude in the 80 Study Regions
(A) Regional plant diversity, i.e., the number of vascular plant species (log10-scale) in equal area grids of ≈ 12,100 km².
(B) Mean local plant diversity, i.e., the effective number of plant species in each network (e to the power of Shannon diversity of plant species interaction frequencies), averaged over multiple networks from the same region. Red triangles indicate regions with pollination networks, blue triangles those with seed dispersal networks. Note that the estimates of regional plant species richness are likely to underestimate the latitudinal gradient in the diversity of animal-pollinated and animal-dispersed plants: while the proportions of animal-pollinated and animal-dispersed plants increase in the tropics [29], we relied on overall estimates of vascular plant species richness for this analysis. Regional and average local plant diversity were not correlated (n = 78, r = 0.077, p = 0.505).
Dissertation J Fründ 115
Figure S4, Related to Table 1. Spatial Autocorrelation in the Residuals of Minimal Adequate Linear Models
(A) Absolute latitude, (B) past climate stability, (C) contemporary climate, (D) regional plant diversity, and (E) local plant diversity. Minimal adequate linear models are provided in Table 1.
Note that similarity in the residuals of all models did not decrease with increasing distance of discrete distance classes of 500 km, i.e., spatial autocorrelation was negligibly small in all minimal adequate models. Red dots indicate Moran's I similarities significantly different from 0 (two-sided permutation test, p < 0.025).
Dissertation J Fründ 116
Table S1, Related to Figure 1. Detailed Information about Location and Sampling Intensity for Each of the 80 Sampling Regions
For each sampling region, we provide the name of the data holders, network type (pollination or seed dispersal), latitude and longitude [decimal degrees], country, altitude [m above sea level], glaciation at last glacial maximum (21,000 years ago), predominant habitat type (forest or non-forest), completeness of sampling (full species communities or restricted to specific plant and/or animal families), sampling focus (plant or animal) and sampling design (sampling time representative for species abundance or standardized per species). We further provide the number of networks per region and means across all networks from a region for sampling duration [observation days], number of animal and plant species, number of observed interaction events as well as network specialization ΔH2'.
Dissertation J Fründ 117
Table S1 P_1Abrahamczykpollination-17.0-65.1Bolivia302noforestfullplantstandardized P_2Abrahamczykpollination-17.6-63.4Bolivia411noforestfullplantstandardized P_3Abrahamczykpollination-18.7-63.2Bolivia434noforestfullplantstandardized P_4Abrahamczykpollination-21.6-62.5Bolivia268noforestfullplantstandardized P_5Alarconpollination39.0-107.0USA3420yesnon-forestfullplantrepresentative P_6Alarconpollination34.2-117.0USA2300nonon-forestfullplantrepresentative P_7Albrechtpollination46.49.9Switzerland1984yesnon-forestfullplantrepresentative P_8Bommarcopollination59.817.5Sweden20yesnon-forestfullplantrepresentative P_9Barrettpollination46.6-66.0Canada120yesforestfullplantstandardized P_10Bauerpollination45.0-109.4USA3050yesnon-forestrestrictedplantrepresentative P_11Bazarianpollination-23.0-48.1Brasil700noforestfullplantstandardized P_12Bluethgenpollination53.113.9Germany30yesnon-forestfullplantrepresentative P_13Bluethgenpollination51.210.4Germany350nonon-forestfullplantrepresentative P_14Bluethgenpollination48.49.5Germany800nonon-forestfullplantrepresentative P_15Dalsgaardpollination19.5-105.1Mexico265noforestrestrictedplantrepresentative P_16Dalsgaardpollination10.7-61.3Trinidad185noforestrestrictedplantrepresentative P_17Dalsgaardpollination9.5-83.5Costa Rica3150noforestrestrictedplantrepresentative P_18Dalsgaardpollination5.9-73.4Colombia2400noforestrestrictedplantrepresentative P_19Dalsgaardpollination4.5-73.9Colombia2475noforestrestrictedplantrepresentative P_20Dalsgaardpollination0.0-78.8Ecuador1650noforestrestrictedplantrepresentative P_21Dalsgaardpollination-8.6-38.6Brasil321noforestrestrictedplantrepresentative P_22Dalsgaardpollination-13.1-41.6Brasil940noforestrestrictedplantrepresentative P_23Dalsgaardpollination-20.0-43.9Brasil1325noforestrestrictedplantrepresentative P_24Dalsgaardpollination-20.8-42.9Brasil785noforestrestrictedplantrepresentative P_25Dalsgaardpollination-23.5-45.9Brasil850noforestrestrictedplantrepresentative P_26Dickspollination52.61.3UK20nonon-forestfullplantrepresentative P_27Dworschakpollination5.0117.8Malaysia100noforestfullplantrepresentative P_28Elberlingpollination68.318.5Sweden1000yesnon-forestfullplantrepresentative P_29Fruendpollination49.910.2Germany308nonon-forestfullplantrepresentative P_30Gotliebpollination30.835.3Israel-155nonon-forestfullplantrepresentative P_31Hagenpollination0.334.9Kenya1600noforestfullplantrepresentative P_32Harterpollination-29.5-50.2Brasil750noforestfullplantrepresentative P_33Holzschuhpollination51.59.9Germany150nonon-forestfullplantrepresentative P_34Inouyepollination-36.4148.3Australia1990nonon-forestfullplantrepresentative P_35Kaepylaepollination60.222.0Finland25yesnon-forestrestrictedplantrepresentative P_36Kaiser-Bunburypollination-4.755.5Seychelles460nonon-forestfullplantstandardized P_37Kaiser-Bunburypollination-20.457.5Mauritius650nonon-forestfullplantstandardized P_38Katopollination35.3135.9Japan350noforestfullplantrepresentative P_39Kevanpollination81.8-71.3Canada300yesnon-forestfullplantrepresentative P_40Koenigerpollination7.080.0Sri Lanka50nonon-forestrestrictedanimalrepresentative P_41Memmottpollination51.4-2.6UK65nonon-forestfullplantrepresentative P_42Mosquinpollination75.0-115.0Canada100nonon-forestfullplantrepresentative P_43Mottenpollination36.0-78.9USA100noforestfullplantrepresentative P_44Ollertonpollination-29.630.1South Africa1200nonon-forestrestrictedplantrepresentative P_45Poursinpollination44.61.1France230nonon-forestrestrictedanimalrepresentative P_46Queirozpollination-25.2-48.8Brasil150noforestrestrictedplantrepresentative P_47Readerpollination43.9-80.4Canada490yesnon-forestrestrictedplantstandardized P_48Schemskepollination40.1-88.2USA220yesforestfullplantrepresentative P_49Smallpollination45.4-75.5Canada70yesnon-forestfullplantrepresentative P_50Ssymankpollination53.313.7Germany75yesnon-forestrestrictedplantrepresentative P_51Ssymankpollination50.67.1Germany160nonon-forestrestrictedplantrepresentative P_52Stilespollination10.4-84.0Costa Rica50noforestrestrictedplantrepresentative P_53Varassinpollination-20.0-40.5Brasil700noforestrestrictedplantrepresentative P_54Vazquezpollination-41.1-71.5Argentina969yesforestfullplantrepresentative P_55Wattspollination-12.9-69.4Peru260noforestfullplantstandardized P_56Wattspollination-13.2-72.2Peru3526noforestrestrictedplantrepresentative P_57Williamspollination38.7-122.2USA203nonon-forestrestrictedplantrepresentative P_58Williamspollination7.2-58.6Guyana35noforestfullplantrepresentative
DesignGlaciation LGMHabitatCompletenessRegion_IDFocusData holderNetwork typeCountryLatitudeLongitudeAltitude
Dissertation J Fründ 118
Table S1 (continued) P_1Abrahamczyk P_2Abrahamczyk P_3Abrahamczyk P_4Abrahamczyk P_5Alarcon P_6Alarcon P_7Albrecht P_8Bommarco P_9Barrett P_10Bauer P_11Bazarian P_12Bluethgen P_13Bluethgen P_14Bluethgen P_15Dalsgaard P_16Dalsgaard P_17Dalsgaard P_18Dalsgaard P_19Dalsgaard P_20Dalsgaard P_21Dalsgaard P_22Dalsgaard P_23Dalsgaard P_24Dalsgaard P_25Dalsgaard P_26Dicks P_27Dworschak P_28Elberling P_29Fruend P_30Gotlieb P_31Hagen P_32Harter P_33Holzschuh P_34Inouye P_35Kaepylae P_36Kaiser-Bunbury P_37Kaiser-Bunbury P_38Kato P_39Kevan P_40Koeniger P_41Memmott P_42Mosquin P_43Motten P_44Ollerton P_45Poursin P_46Queiroz P_47Reader P_48Schemske P_49Small P_50Ssymank P_51Ssymank P_52Stiles P_53Varassin P_54Vazquez P_55Watts P_56Watts P_57Williams P_58Williams Region_IDData holder 242616720.24S. Abrahamczyk, J. Kluge, Y. Gareca, S. Reichle, M. Kessler,Plos One6, e27115 (2011). 24278780.46S. Abrahamczyk, J. Kluge, Y. Gareca, S. Reichle, M. Kessler, Plos One6, e27115 (2011). 1427111430.25S. Abrahamczyk, J. Kluge, Y. Gareca, S. Reichle, M. Kessler,Plos One6, e27115 (2011). 1445163460.30S. Abrahamczyk, J. Kluge, Y. Gareca, S. Reichle, M. Kessler,Plos One6, e27115 (2011). 1351364121110.32R. Alarcón, thesis, University of California, Riverside, USA (2004). 1391273817110.40R. Alarcón, N. M. Waser, J. Ollerton,Oikos 117, 1796 (2008). 5722712810.38M. Albrecht, M. Riesen, B. Schmid, Oikos119, 1610 (2010). 4742322960.36C. Westphal et al., Ecol. Monogr.78, 653 (2008). 1150102125500.46S. C. H. Barrett, K. Helenurm, Can. J. Bot.65, 2036 (1986). 1611174530.25P. J. Bauer,Am. J. Bot.70, 134 (1983). 9449194520.54S. V. Bazarian, thesis, Universidade de São Paulo, São Paulo, Brazil (2010). 2313781570.39Data collected by C. N. Weiner, M. Werner, and N. Blüthgen (coauthors) in 2008 in the Schorfheide Biosphere Reserve, Germany. 3613992240.44Data collected by C. N. Weiner, M. Werner, and N. Blüthgen (coauthors) in 2008 in the Hainich National Park, Germany. 66147122690.42C. N. Weiner, M. Werner, K. E. Linsenmair, N. Blüthgen,Basic Appl. Ecol.12, 292 (2011). 136551561330.39M. C. Arizmendi, J. F. Ornelas,Biotropica22, 172 (1990). 136595714170.36B. K. Snow, D. W. Snow,J. Anim. Ecol. 41, 471 (1972). 185251370.60L. L. Wolf, F. G. Stiles, F. R. Hainsworth, J. Anim. Ecol.45, 349 (1976). 12512223430.49D. W. Snow, B. K. Snow,Bull. Br. Mus. (Nat. Hist.) Zool.38, 105 (1980). 1239133040.37D. W. Snow, B. K. Snow,Bull. Br. Mus. (Nat. Hist.) Zool.38, 105 (1980). 170196521620.33B. A. Walther, H. Brieschke, Int. J. Ornithol.4, 115 (2001). 1365472640.18F. C. Leal, A. V. Lopes, I. C. Machado,Rev. Bras. Bot.29, 379 (2006). 136572825190.40C. G. Machado,Zoologia26, 255 (2009). 13656107750.30M. F. Vasconcelos, J. A. Lombardi,Ararajuba 7, 71 (1999). 13658141780.34C. R. M. Abreu, M. F. Vieira, Lundiana5, 129 (2004). 1376252500.31D. W. Snow, B. K. Snow,El Hornero 12, 286 (1986). 212491720650.41L. V. Dicks, S. A. Corbet, R. F. Pywell,J. Anim. Ecol.71, 32 (2002). 190304317020.47K. Dworschak, N. Blüthgen, Ecol. Entomol. 35, 216 (2010). 1120118233830.21H. Elberling, J. M. Olesen, Ecography22, 314 (1999). 2112311760.31J. Fründ, K. E. Linsenmair, N. Blüthgen,Oikos119, 1581 (2010). 4901662050.31Data collected by A. Gotlieb (coauthor) in 2009 in the Rift Valley, Israel. 3365552810280.23M. Hagen, M. Kraemer, Biol. Conserv.143, 1654 (2010). 136518518450010.31B. Harter, thesis, University of Tübingen, Germany (1999). 46089700.33Data collected by A. Holzschuh, C.F. Dormann, T. Tscharntke (coauthors) in 2006 in the surroundings of Göttingen, Germany. 182834114590.52D. W. Inouye, G. H. Pyke, Aust. J. Ecol. 13, 191 (1988). 11505342110.56M. Käpylä, Biological Research Reports University of Jyväskylä5, 3 (1978). 3240561916070.27C. N. Kaiser-Bunbury, T. Valentin, J. Mougal, D. Matatiken, J. Ghazoul, J. Ecol.99, 202 (2011). 12101357339610.25C. N. Kaiser-Bunbury, J. Memmott, C. B. Müller,Perspect. Plant Ecol. Evol. Syst.11, 241 (2009). 11806799123920.38M. Kato, T. Makutani, T. Inoue, T. Itino,Contribution from the Biological Laboratory, Kyoto University27, 309 (1990). 1731143025230.45P. G. Kevan, thesis, University of Alberta, Canada (1970). 1604351420.15N. Koeniger, G. Vorwohl,J. Apic. Res.18, 95 (1979). 130792521830.21J. Memmott,Ecol. Lett.2, 276 (1999). 11318111340.36T. Mosquin, J. E. H. Martin,Can. Field-Nat.81, 201 (1967). 160441322250.40A. F. Motten,Ecol. Monogr.56, 21 (1986). 1635695940.37J. Ollerton, S. D. Johnson, L. Cranmer, S. Kellie, Ann. Bot. 92, 807 (2003). 19011123210.45J. M. Poursin,Apidologie13, 227 (1982). 13658122050.42V. Q. Piacentini, I. G. Varassin,J. Trop. Ecol.23, 663 (2007). 11201445960.38R. J. Reader,Can. J. Bot. 53, 1300 (1975). 1903272990.26D. W. Schemskeet al., Ecology59, 351 (1978). 16034139920.49E. Small,Can. Field-Nat.90, 22 (1976). 112052443820.39A. Ssymank,Volucella6, 81 (2002). 1180758848370.44A. Ssymank,Schriftenreihe für Landschaftpflege und Naturschutz 64, 1 (2001). 13659910260.22F. G. Stiles,Ecology56, 285 (1975). 136514201400.13I. G. Varassin, M. Sazima, Bol. Mus. Biol. Mello Leitão Nova Sér11/12, 57 (2000). 81502996620.67D. P. Vázquez, D. Simberloff,Am. Nat.159, 606 (2002). 115875010.22S. Watts, thesis, University of Northampton, UK (2008). 9137231760.25S. Watts, thesis, University of Northampton, UK (2008). 518069343990.38N. M. Williams, D. Cariveau, R. Winfree, C. Kremen, Basic Appl. Ecol.12, 332 (2011). 1151791860.52N. H. Williams, C. H. Dodson, Evolution 26, 84 (1972).
mean # plantsmean # interactionsmean duration [days]# networksmean # animalsH2'Further information
Dissertation J Fründ 119
Table S1 (continued) DesignGlaciation LGMHabitatCompletenessRegion_IDFocusData holderNetwork typeCountryLatitudeLongitudeAltitude S_1Bairdseed dispersal40.3-74.7USA20noforestfullplantrepresentative S_2Beehlerseed dispersal-7.3146.7P. New Guinea1430noforestrestrictedplantstandardized S_3Carloseed dispersal18.3-66.6Puerto Rico408noforestfullanimalrepresentative S_4Dehlingseed dispersal-13.1-71.6Peru2200noforestfullplantrepresentative S_5Engelseed dispersal-4.239.4Kenya190noforestfullanimalrepresentative S_6Faria/Galettiseed dispersal-22.8-47.1Brasil650noforestfullanimal/plantrepresentative S_7Frostseed dispersal-29.031.8South Africa20noforestfullplantrepresentative S_8Gorchovseed dispersal-4.9-73.8Peru120noforestfullanimalrepresentative S_9Hovestadtseed dispersal9.0-3.6Ivory Coast240nonon-forestfullplantrepresentative S_10Jordanoseed dispersal37.6-2.5Spain1615noforestfullplantrepresentative S_11Kantakseed dispersal18.5-89.5Mexico280noforestfullplantstandardized S_12Passosseed dispersal-24.3-48.4Brasil615noforestrestrictedanimalrepresentative S_13Pedroseed dispersal-19.2-48.4Brasil800noforestrestrictedanimalrepresentative S_14Poulinseed dispersal9.2-79.7Panama150noforestfullanimalrepresentative S_15Schleuningseed dispersal50.38.7Germany208nonon-forestfullplantrepresentative S_16Schleuningseed dispersal0.434.9Kenya1600noforestfullplantstandardized S_17Silveiraseed dispersal-22.4-47.0Brasil610noforestrestrictedanimalrepresentative S_18Snowseed dispersal51.8-0.8UK100nonon-forestrestrictedplantrepresentative S_19Snowseed dispersal10.7-61.2Trinidad550noforestfullplantrepresentative S_20Sorensenseed dispersal51.8-1.3UK120noforestfullplantrepresentative S_21Stiebelseed dispersal51.29.0Germany300noforestfullplantrepresentative S_22Stiles/Lopezseed dispersal10.4-84.0Costa Rica50noforestrestrictedplantrepresentative
Dissertation J Fründ 120
Table S1 (continued) Region_IDData holder S_1Baird S_2Beehler S_3Carlo S_4Dehling S_5Engel S_6Faria/Galetti S_7Frost S_8Gorchov S_9Hovestadt S_10Jordano S_11Kantak S_12Passos S_13Pedro S_14Poulin S_15Schleuning S_16Schleuning S_17Silveira S_18Snow S_19Snow S_20Sorensen S_21Stiebel S_22Stiles/Lopez mean # plantsmean # interactionsmean duration [days]# networksmean # animalsH2'Further information 11802176550.44J. W. Baird,Wilson Bull. 92, 63 (1980). 124993111890.23B. Beehler,Auk100, 1 (1983). 44816262370.27T. A. Carlo, J. A. Collazo, M. J. Groom, Oecologia134, 119 (2003). 21647349240.29Data collected by D.M. Dehling (coauthor) between December 2009 and February 2010 in the Manú Biosphere Reserve, Peru. 13653321937300.35T. R. Engel, thesis, University of Bayreuth, Germany (2000). 236519262720.23D. M. Faria, thesis, Universidade Estadual de Campinas, São Paulo, Brazil (1996); M. Galetti, M. A. Pizo,Ararajuba 4, 71 (1996). 1365101635540.24P. G. H. Frost, in Acta XVII Congressus Internationalis Ornithologici, R. Noring, Ed. (Berlin, 1980), pp. 1179-1184. 1365189111860.34D. L. Gorchov, F. Cornejo, C. F. Ascorra, M. Jaramillo, Oikos74, 235 (1995). 13654834175750.17T. Hovestadt, thesis, University of Würzburg, Germany (1997). 1365332570100.34P. Jordano,Ardeola32, 69 (1985). 18027555490.31G. E. Kantak, Auk96, 183 (1979). 13656231010.25F. C. Passos, W. R. Silva, W. A. Pedro, M. R. Bonin, Rev. Bras. Zool.20, 511 (2003). 13657131080.35W. A. Pedro, thesis, Universidade de Campinas, Campinas, Brazil (1992). 18420174920.15B. Poulin, S. J. Wright, G. Lefebvre, O. Calderón, J. Trop. Ecol.15, 213. 86015102230.48M. Plein, thesis, University of Mainz, Germany (2011). 190883334470.25M. Schleuning et al., Ecology92, 26 (2011). 1365671820.13M. Silveira, thesis, Universidade Estadual de São Paulo, Rio Claro, Brazil (2006). 13001929199460.30B. K. Snow, D. W. Snow,Birds and Berries (T & AD Poyser, Calton, England, 1988). 160145021440.25B. K. Snow, D. W. Snow,J. Anim. Ecol. 41, 471 (1972). 1220141174340.46A. E. Sorensen, Oecologia50, 242 (1981). 1365302963600.39H. Stiebel, F. Bairlein,Vogelwarte46, 1 (2008). 230321227980.20F. G. Stiles,Brenesia15, 151 (1979); J. E. Lopez, C. Vaughan,Revista de Biología Tropical55, 301 (2007).
Dissertation J Fründ 121
Table S2, Related to Figure 2. Minimal Adequate Linear Models of the Effects of Multiple Predictor Variables on Network Specialization (ΔH2') in 80 Study Regions
(A) All minimal adequate linear models with ΔAICc < 2
Predictor β t p
Best model, R² = 0.32, p < 0.001
Network type (pollination) 0.058 2.46 0.016
Growing degree days –0.594 –5.29 < 0.001
Habitat type (forest) 0.065 2.47 0.016
Alternative model, ΔAICc = 0.94, R² = 0.34, p < 0.001
Network type (pollination) 0.068 2.71 0.008
Growing degree days –0.563 –4.89 < 0.001
Habitat type (forest) 0.071 2.65 0.010
Taxonomic focus (full) 0.028 1.15 0.253
Alternative model, ΔAICc = 1.62, R² = 0.35, p < 0.001
Network type (pollination) 0.127 2.66 0.010
Growing degree days –0.552 –4.74 < 0.001
Habitat type (forest) 0.067 2.50 0.015
Climate-change velocity 0.331 1.69 0.096
Network type x Climate-change velocity –0.351 –1.60 0.113 Alternative model, ΔAICc = 1.63, R² = 0.33, p < 0.001
Network type (pollination) 0.052 2.09 0.040
Growing degree days –0.572 –4.96 < 0.001
Habitat type (forest) 0.068 2.56 0.013
Observation time span –0.087 –0.82 0.418
Dissertation J Fründ 122
Predictors were past climate stability (glaciation during LGM, climate-change velocity), contemporary climate (growing degree days), and potential confounding factors (time span of observation, habitat type, taxonomic sampling focus); network type (pollination, seed dispersal) was included in all models. Based on the results of univariate models, we included the interaction term between climate-change velocity and network type (pollination, seed dispersal) in all models with climate-change velocity.
(B) Akaike weights for all predictor variables across all 63 model combinations.
Predictor variable Akaike weight Growing degree days 0.999
Habitat type 0.898
Taxonomic focus 0.343
Climate-change velocity 0.323
Sampling period 0.308
Glaciated during LGM 0.245
We fitted linear models for all combinations of predictor variables (n = 63 models) and calculated the Akaike weights for each fitted model. The Akaike weight gives the likelihood that a model is the best available model, and thus the summed Akaike weight for each predictor variable measures the relative importance of each variable in contributing to the best model.
Note that none of the potentially confounding variables (habitat type, taxonomic focus, sampling period) significantly affected ΔH2' in univariate models (p > 0.05 in all cases).
Dissertation J Fründ 123
Table S3, Related to Table 1. Correlations between Different Specialization Metrics and Sampling Effort and Network Size
(A) Sampling effort (B) Network size
r p r p
Specialization ΔH2' –0.015 0.898 0.075 0.507
Specialization H2' –0.361 0.001 0.093 0.413
Connectance 0.027 0.810 –0.700 <0.001
Unweighted Generality 0.446 <0.001 0.088 0.438
Weighted Generality 0.313 0.005 0.076 0.501
Plant specialization di' –0.237 0.035 0.320 0.004 Animal specialization dj' –0.268 0.017 0.167 0.138 (A) Sampling effort is estimated by the number of observed interactions events (log10-scale), and (B) network size equals the sum of plant and animal species in a network (log10-scale).
Pearson correlation coefficients r and p-values are given; significant correlations are printed bold. Pearson correlations r were calculated with region as the unit of replication (n = 80 in all cases). ΔH2' is the only index that is neither related to sampling effort nor to network size and was therefore the preferred metric in the main manuscript.
Dissertation J Fründ 124