Electronic Supplementary Material (ESM) for Davis et al. “The utilization of phylogeny in assessing species’ phenological response to global climate change”
Figure S1. Composite phylogenies of 167 and 323 flowering plant species from Chinnor, England (A) and Concord, USA (B) depicted in Figure 2 of the main text. Red and blue dots at nodes represent less and more phenologically responsive clades, respectively, based on their ability to track their flowering times to seasonal temperature variation. Solid colored dots indicate those clades that have demonstrated a significant tracking response. A null distribution of mean flowering time tracking for each node in the phylogeny was generated by 9999 randomizations (‘node.mn’ in Phylocom). Observed clade values that ranked ≤ 250 (p ≤ 0.05) indicated significantly strong trackers while values that ranked ≥ 9750 (p ≤ 0.05) indicated significantly weak trackers. Open dots indicate marginally significant clades (p ≤ 0.10). Each of the clades exhibiting these tracking responses is further highlighted in pink and blue and referenced numerically to its clade name. Subclades of interest are labeled with the more familiar, and more inclusive, clade to which they belong followed by pro parte (p.p.).
Figure S2. Composite phylogeny of 100 European bird species depicted in Figure 3 of the main text. Solid red dots indicate clades that have declined significantly from 1990-2000 (results were similar for the earlier interval 1970-1990); solid blue dots indicate clades that have increased significantly. A null distribution of mean change in abundance for each clade in the phylogeny was generated by 9999 randomizations (‘node.mn’ in Phylocom). Observed clade values that ranked ≤ 250 (p ≤ 0.05) were considered to have significantly declined in abundance while values that ranked ≥ 9750 (p ≤ 0.05) were considered to have significantly increased in abundance. Open dots indicate marginally significant clades (p ≤ 0.10). Each of these clades is further highlighted in pink or blue and referenced numerically to its clade name. Subclades of interest are labeled with the more familiar, and more inclusive, clade to which they belong followed by pro parte (p.p.).
Traits / Poales / n / Mean Obs. Rankbldj / p-valueConcord, USA
Non-native status / - / 587 / 1 / 0.0002
Invasive status / - / 587 / 200 / 0.0400
Flowering time tracking / - / 167 / 2884 / 0.5768
Flowering time shift (1900-2006) / - / 342 / 2143 / 0.4286
Chinnor, England
Non-native status / Poales / 377 / 873 / 0.1746
Invasive status / Poales / 377 / 1584 / 0.3168
Non-native status / Poales
excluded / 323 / 94 / 0.0188
Invasive status / Poales
excluded / 323 / 4446 / 0.8892
Flowering time tracking / Poales / 377 / 972 / 0.1943
Flowering time shift (1954-2000) / Poales / 377 / 1783 / 0.3567
Flowering time tracking / Poales
excluded / 323 / 154 / 0.0307
Flowering time shift (1954-2000) / Poales
excluded / 323 / 4305 / 0.8609
Table S1. Phylogenetic signal of non-native species status and phenological response traits when accounting for phylogenetic uncertainty. Each trait was analyzed across a set of 50 trees with polytomies resolved randomly. Branch lengths were adjusted for estimated divergence time. Mean obs. rank of the 50 iterations is presented in the table. A high mean obs. rank indicates trait overdispersion. A low mean obs. rank indicates trait clustering (i.e., phylogenetic signal). Significance of observed rank based on a two-tail test: significant obs. rank < 250, > 9750 = P, 0.05, marginally significant obs. rank < 500, > 9500 = P, 0.10.
Traits / n / Obs. Rankbldj / p-value / Obs. RankBL1 / p-valueNon-native status / 377 / 1266 / 0.2532 / 440 / 0.0880
Invasive status / 377 / 4163 / 0.2532 / 9906 / 0.0188
Flowering time tracking / 377 / 3890 / 0.7780 / 56 / 0.0112
Flowering time shift (1954-2000) / 377 / 4054 / 0.8108 / 8150 / 0.3700
Table S2. Phylogenetic signal of non-native species status and phenological response traits for Chinnor, England with the Poales clade included. Phylogenetic signal was tested by comparing the rank of the observed mean standard deviation (obs. rank) calculated across all nodes in the community phylogeny to a null distribution of mean standard deviations based on 9999 iterations of trait values arranged randomly across the phylogeny tips. These analyses were conducted with branch lengths adjusted for estimated divergence times (obs. rankbldj) and with branch lengths set to 1 (obs. rankBL1). A high obs. rank indicates trait overdispersion. A low mean obs. rank indicates trait clustering. Significance of observed rank based on a two-tail test: significant obs. rank < 250, > 9750 = P, 0.05, marginally significant obs. rank < 500, > 9500 = P, 0.10.
Traits / Non-native vs. Native / Invasive vs. Native / Non-invasive non-native vs. Nativen1 / n2 / n3 / estimate / t / p-value / estimate / t / p-value / estimate / t / p-value
Branch lengths scaled to approximate time
Flowering time shift (1954-2000) / 322 / 44 / 11 / -0.45 / -1.09 / 0.2804 / -5.40 / -5.93 / 0.0000 / 0.40 / 0.92 / 0.3616
Flowering time tracking / 322 / 44 / 11 / -0.14 / -8.94 / 0.0000 / 0.00 / -0.12 / 0.9011 / -0.16 / -9.93 / 0.0000
Branch lengths set to one
Flowering time shift (1954-2000) / 322 / 44 / 11 / 0.48 / 1.80 / 0.2980 / -5.78 / -6.83 / 0.0000 / 2.33 / 4.67 / 0.0000
Flowering time tracking / 322 / 44 / 11 / -0.14 / -7.04 / 0.0000 / -0.06 / -1.55 / 0.1269 / -0.16 / -7.51 / 0.0000
Table S3. Correlations of phenological response traits for non-native species from Chinnor, England with the Poales clade included. Analyses were conducted with branch lengths adjusted for time and branch lengths set to one. Trait correlations of these phenological response traits (i.e., flowering time shift and flowering time tracking) and non-native/native classes were tested using generalized estimating equations (see main text). n = species sample size within each class: 1) natives, 2) non-native non-invasives and 3) invasives
Traits / Non-native vs. Native / Invasive vs. Native / Non-invasive non-native vs. NativeConcord, Massachusetts / Poales / n1 / n2 / n3 / estimate / t / p-value / estimate / t / p-value / estimate / t / p-value
Flowering time shift (1900-2006) / - / 259 / 72 / 11 / 0.50 / 0.56 / 0.4167 / 13.55 / 6.26 / 0.0000 / -1.08 / -1.18 / 0.3438
Flowering time tracking / - / 134 / 28 / 5 / -0.11 / -4.22 / 0.0003 / -0.19 / -2.94 / 0.0138 / -0.10 / -3.82 / 0.0010
Chinnor, England
Flowering time shift (1954-2000) / Poales / 322 / 44 / 11 / -0.48 / -1.15 / 0.3139 / -5.57 / -6.16 / 0.0000 / 0.42 / 0.95 / 0.2595
Flowering time tracking / Poales / 322 / 44 / 11 / -0.13 / -8.13 / 0.0000 / 0.00 / 0.12 / 0.7841 / -0.15 / -9.15 / 0.0000
Flowering time shift (1954-2000) / Poales
excluded / 271 / 41 / 11 / -0.49 / -0.97 / 0.3837 / -5.45 / -5.81 / 0.0000 / 0.81 / 1.47 / 0.2305
Flowering time tracking / Poales
excluded / 271 / 41 / 11 / -0.08 / -4.27 / 0.0019 / 0.02 / 0.53 / 0.5964 / -0.11 / -5.25 / 0.0002
Table S4. Correlations of phenological response traits for non-native species accounting for phylogenetic uncertainty. Each analysis was run using 50 different trees with polytomies resolved randomly. Branch lengths were adjusted for time after resolution. Trait correlations of these phenological response traits (i.e., flowering time shift and flowering time tracking) and non-native/native classes were tested using generalized estimating equations (see main text). The mean of the 50 iterations are presented in the table. n = species sample size within each class: 1) natives, 2) non-native non-invasives and 3) invasives.
Traits / Non-native vs. Native / Invasive vs. Native / Non-invasive non-native vs. NativeConcord, USA / n1 / n2 / n3 / estimate / t / p-value / estimate / t / p-value / estimate / t / p-value
Flowering time shift (1900-2006) / 245 / 65 / 11 / 1.42 / 1.64 / 0.11 / 9.57 / 5.17 / <.0001 / 0.004 / 0.004 / 0.99
Flowering time tracking / 126 / 25 / 5 / -0.12 / -3.68 / 0.0001 / -0.080 / -1.02 / 0.32 / -0.12 / -3.69 / 0.001
Chinnor, England
Flowering time shift (1954-2000) / 271 / 41 / 11 / 0.56 / 1.15 / 0.2559 / -5.74 / -6.55 / <.0001 / 2.55 / 4.78 / <.0001
Flowering time tracking / 271 / 41 / 11 / -0.13 / -6.24 / <.0001 / -0.05 / -1.44 / 0.1565 / -0.15 / -6.67 / <.0001
Table S5. Phylogenetic correlations of two phenological response traits across non-native and native species’ classes in Concord, USA and Chinnor, England. Phylogentic branch lengths have been adjusted to one. Trait correlations of these phenological response traits (i.e., flowering time shift and flowering time tracking) and non-native/native classes were tested using generalized estimating equations (see main text). n = species sample size within each class: 1) natives, 2) non-native non-invasives and 3) invasives.
Traits / Change in abundance 1990-2000 / Change in abundance 1970-1990n / estimate / t / p-value / estimate / t / p-value
Habitat specificity / 38 / -1.28 / -7.26 / 0.0012 / -0.45 / -1.91 / 0.1208
Thermal maximum / 38 / 0.39 / 4.98 / 0.0055 / 0.02 / 0.22 / 0.8347
Natal dispersal / 45 / -0.18 / -0.82 / 0.4445 / 0.00 / 0.00 / 0.9995
Change in migration date / 98 / -2.81 / -12.59 / 0.0000 / -1.18 / -4.54 / 0.0007
Migration distance / 98 / -0.77 / -5.85 / 0.0001 / -0.60 / -4.43 / 0.0009
Overwinter in Africa / 98 / -0.65 / -4.94 / 0.0003 / -0.93 / -6.93 / 0.0000
Number of broods / 98 / 0.29 / 3.66 / 0.0033 / 0.13 / 1.65 / 0.1248
Farmland habitat / 98 / -0.43 / -2.95 / 0.0122 / -0.70 / -4.62 / 0.0006
Northernmost latitude / 98 / 0.03 / 1.58 / 0.1404 / -0.05 / -3.16 / 0.0085
Sexual dichromatism / 98 / -0.17 / -1.27 / 0.2276 / -0.10 / -0.68 / 0.5099
European population size / 98 / -0.10 / -1.09 / 0.2954 / 0.11 / 1.06 / 0.3099
Body mass / 98 / 0.07 / 0.33 / 0.7451 / 0.86 / 4.55 / 0.0007
Change in abundance (1970-1990) / 96 / 0.23 / 5.88 / 0.0001 / - / - / -
Table S6. Univariate regression analyses of trait correlations with change in abundance in European birds. The individual regression between each trait and change in abundance (calculated over two time periods: 1970-1990, 1990-2000) was tested using generalized estimating equations. Estimates indicate the direction and magnitude of the correlation. n = species sample size.
Traits / Change in abundance 1990-2000 / Change in abundance 1970-1990n / estimate / t / p-value / estimate / t / p-value
Habitat specificity / 38 / -0.59 / -3.71 / 0.0989 / -0.41 / -1.57 / 0.2326
Thermal maximum / 38 / 0.29 / 4.66 / 0.0717 / -0.03 / -0.32 / 0.7757
Change in migration date / 38 / -2.97 / -12.53 / 0.0166 / -0.47 / -1.19 / 0.3356
European population size / 38 / 0.07 / 1.51 / 0.3081 / - / - / -
Table S7. Multivariate regression analyses of trait correlations with abundance change in European birds using generalized estimating equations. All traits listed were included as co-variates in the regression model. Estimates indicate the direction and magnitude of the correlation. n = species sample size.
Chinnor, England / Concord, USATaxon / n / mean / SD / rank / Taxon / n / mean / SD / rank
Rubiaceae / 5 / 1.25 / 0.10 / 1100 / Rubiaceae / 4 / 1.42 / 0.04 / 786
Rosaceae p.p. (Prunus, Malus, Sorbus) / 7 / 1.55 / 0.22 / 8 / Rosaceae p.p. (Prunus, Amelanchier) / 5 / 1.36 / 0.07 / 857
Rosaceae p.p. (Geum, Rosa) / 10 / 0.58 / 0.00 / 7943 / Rosaceae p.p. (Rosa, Potentilla) / 4 / 0.77 / 0.05 / 8119
Ranunculaceae p.p. (Ranuculus) / 5 / 1.29 / 0.22 / 657 / Ranunculaceae p.p. (Ranunculus) / 3 / 1.55 / 0.12 / 405
Ranunculaceae p.p. (Clematis, Ranuculus) / 7 / 1.30 / 0.11 / 635 / Ranuculaceae (Clematis, Ranuculus) / 9 / 0.87 / 0.01 / 7715
Onagraceae / 6 / 0.75 / 0.31 / 6134 / Onagraceae / 4 / 0.77 / 0.28 / 8324
Lamiales / 40 / 0.55 / 0.07 / 9711 / Lamiales / 14 / 0.79 / 0.01 / 9233
Fabaceae p.p. (Trifolium) / 5 / 0.84 / 0.23 / 5102 / Fabaceae p.p. (Trifolium) / 5 / 1.24 / 0.02 / 2527
Campanulaceae / 2 / -0.02 / 0.13 / 9371 / Campanulaceae / 2 / 0.74 / 0.07 / 7644
Asteraceae p.p. (Senecio, Solidago) / 21 / 0.60 / 0.16 / 8941 / Asteraceae p.p. (Senecio, Solidago) / 14 / 0.74 / 0.11 / 800
Apiales / 13 / 0.39 / 0.49 / 9374 / Apiaceae / 4 / 0.84 / 0.23 / 7489
Asterales / 54 / 0.26 / 0.13 / 9996 / Asparagales / 7 / 1.19 / 0.26 / 2715
Betulaceae / 4 / 1.39 / 0.12 / 479 / Papaveraceae / 2 / 1.59 / 0.07 / 743
Brassicaceae / 15 / 1.18 / 0.09 / 334 / Primulaceae / 4 / 1.19 / 0.05 / 3292
Caryophyllaceae p.p. (Cerastium, Stellaria) / 6 / 1.38 / 0.13 / 295 / Ericaceae / 15 / 1.28 / 0.12 / 542
Caryophyllales / 25 / 0.34 / 0.04 / 9965 / Cornus / 3 / 0.78 / 0.45 / 8057
Dipsacales / 10 / 0.93 / 0.20 / 3389 / Plantaginaceae / 5 / 0.80 / 0.21 / 7956
Euphorbiaceae / 4 / 1.34 / 0.22 / 753 / Malvaceae/Brassicaceae / 3 / 0.85 / 0.00 / 7290
Fabaceae p.p. (Vicia, Lathyrus) / 4 / 0.48 / 0.10 / 8354
Orchidaceae / 12 / 0.60 / 0.06 / 8457
Plantaginaceae / 12 / 0.85 / 0.15 / 4759
Sapindaceae / 4 / 1.51 / 0.10 / 176
Saxifragales / 6 / 1.34 / 0.02 / 399
Table S8. Clade means for flowering time tracking in Concord and Chinnor (see also Figure 2 in the main text). Mean values are corrected for phylogenetic relatedness and branch lengths. Means have been standardized by their respective phylogenetically corrected community means: a mean of >1 is inferred to be a good tracker, and a mean of <1 is inferred to be a poor tracker. The standard deviation (SD) of these means is also phylogenetically corrected. Significance of means was tested by comparing the rank of the observed mean (rank) to a null distribution based on 9999 iterations of trait values arranged randomly across the clade tips. A significantly low rank indicates that the clade is composed of good trackers, whereas a significantly high rank indicates that the clade is composed of poor trackers. Significance of observed rank is based on a two-tail test (i.e., significant rank < 250, > 9750 = P, 0.05; marginally significant rank < 500, > 9500 = P, 0.10). n = number of terminal taxa.