Ecological Archives E092-083-A1

Sarah E. Diamond, Alicia M. Frame, Ryan A. Martin, and Lauren B. Buckley. 2011. Species' traits predict phenological responses to climate change in butterflies. Ecology 92:1005–1012.

Appendix A. Supporting results for phenological advancement, phylogenetic signal, and statistical model selection.

 

   FIG. A1. Date of first appearance (days since April 1) as a function of time (year) for each of the 44 butterfly species. Points represent annual dates of first appearance; solid lines are the regressions of date against year.


 

 
   FIG. A2. The ability of life history traits to predict phenological changes. Each point represents a single species in all panels. (a–d) Added variable plots (see Velleman and Welsch 1981; partial residuals plotted against the residuals of each independent variable of interest regressed on all remaining independent variables; regressions of partial residuals on the independent variable residuals are indicated with solid lines) based on a model containing all terms identified as part of the top model subset during the model selection process. Only results for nonsignificant predictors of phenological change are shown here; see Fig. 2 for significant predictors. The dashed line at zero corresponds to the mean (±1 SD) change in date of first appearance per decade for all species (-3.92 ± 2.20 days). Points below the dashed line indicate species with greater phenological advancement (more change) compared to points above the line (less change).

 

TABLE A1. Species list and associated phenological, life history and range size data for the 44 UK butterfly species.

Species No. of
years
of data
Change in first
appearance
per decade
Fa Pa Date of first
appearance
in 1975
Voltinism Over-winter
stage
b
Dispersal ability
(mobility score)
No. larval
host plant
species
Percent national
10-km grid
cells occupied
Latitudinal
extent
Seconds North
latitude
Aglais urticae 33 -1.07 0.23 0.64 44.85 2 4 38 2 97.43 4 0.9
Anthocharis cardamines 33 -6.79 31.33 < 0.01 46.25 1 3 32 29 67.39 4 0.8
Aphantopus hyperantus 33 -4.94 35.00 < 0.01 101.38 1 2 16 11 53.78 4 0.77
Argynnis adippe 33 -5.13 19.49 < 0.01 102.81 1 1 28 5 6.83 3 0.64
Argynnis aglaja 33 -0.88 0.99 0.33 99.58 1 2 29 6 32.7 4 0.84
Argynnis paphia 33 -3.84 14.33 < 0.01 112.00 1 2 31 3 19.13 2 0.67
Boloria euphrosyne 33 -7.06 21.42 < 0.01 62.58 1 2 18 7 19.13 4 0.77
Boloria selene 33 -3.60 9.73 < 0.01 78.90 1 2 19 6 33.09 4 0.8
Callophrys rubi 33 -3.68 12.11 < 0.01 53.06 1 3 14 21 31.17 4 0.8
Celastrina argiolus 33 -4.44 4.59 0.04 81.46 2 3 34 20 36.78 3 0.71
Coenonympha pamphilus 33 -1.11 0.78 0.39 68.58 2 2 17 8 88.26 4 0.8
Coenonympha tullia 32 -7.87 26.28 < 0.01 102.53 1 2 4 6 14.65 3 0.84
Cupido minimus 33 -2.90 4.86 0.04 82.10 1 2 1 2 10.26 4 0.8
Erebia aethiops 32 -0.37 0.14 0.72 119.71 1 2 13 6 12.26 2 0.8
Erynnis tages 33 -3.87 8.59 0.01 55.28 1 2 10 4 30.61 4 0.77
Euphydryas aurinia 30 -5.96 17.55 < 0.01 71.17 1 2 13 7 13.35 4 0.74
Gonepteryx rhamni 33 -5.33 18.68 < 0.01 62.35 1 4 36 4 43.7 3 0.71
Hamearis lucina 33 -5.49 25.78 < 0.01 57.56 1 3 9 4 7.09 3 0.64
Hesperia comma 33 -4.43 17.29 < 0.01 138.91 1 1 15 2 1.39 1 0.58
Hipparchia semele 33 -0.85 0.81 0.38 112.55 1 2 22 11 30.17 4 0.8
Inachis io 33 -6.89 24.87 < 0.01 69.18 2 4 39 3 69.96 4 0.9
Lasiommata megera 33 -1.20 0.56 0.46 94.28 2 2 30 10 64.09 3 0.71
Leptidea sinapis 33 -2.67 1.99 0.17 64.24 1 3 14 8 5.65 2 0.64
Limenitis camilla 33 -4.31 17.59 < 0.01 106.79 2 2 27 1 12.48 2 0.61
Lycaena phlaeas 33 -0.31 0.02 0.88 95.65 2 2 26 5 76.96 4 0.84
Maniola jurtina 33 -3.93 15.71 < 0.01 90.55 1 2 25 12 94.91 4 0.84
Melanargia galathea 33 -5.20 33.95 < 0.01 103.34 1 2 24 6 19.65 3 0.68
Melitaea athalia 29 -1.37 1.30 0.26 70.16 1 2 5 10 0.87 1 0.58
Ochlodes sylvanus 33 -3.08 7.98 0.01 86.96 2 2 20 9 57.7 3 0.71
Pararge aegeria 33 -6.42 14.36 < 0.01 54.65 2 2 or 3 23 13 45 4 0.8
Pieris brassicae 33 -2.53 3.45 0.07 79.07 2 3 41 25 85.96 4 0.9
Pieris napi 33 -3.11 6.58 0.02 70.63 2 3 35 21 88.09 4 0.84
Pieris rapae 33 -3.51 7.77 0.01 77.41 2 3 40 26 100 4 0.84
Plebeius agestis 33 -0.74 0.24 0.63 97.84 2 2 12 6 17.7 2 0.64
Plebeius argus 33 -7.05 9.04 0.01 99.99 2 1 2 16 6.74 2 0.61
Plebeius artaxerxes 30 -6.66 7.71 0.01 95.20 1 2 7 3 3.35 3 0.8
Polygonia c-album 33 -9.61 17.51 < 0.01 77.13 2 4 33 8 40.26 3 0.77
Polyommatus bellargus 33 -4.20 8.04 0.01 99.57 2 2 8 2 4.48 1 0.58
Polyommatus coridon 33 -2.86 4.96 0.03 119.37 1 1 11 8 9.61 2 0.61
Polyommatus icarus 33 -3.71 9.83 < 0.01 94.66 2 2 21 11 86.96 4 0.84
Pyrgus malvae 33 -3.67 7.77 0.01 51.60 1 3 6 11 20.87 2 0.64
Pyronia tithonus 33 -2.98 13.75 < 0.01 110.22 1 2 21 11 49.7 3 0.68
Thymelicus acteon 27 -4.49 3.82 0.06 111.40 2 2 5 2 0.61 1 0.51
Thymelicus sylvestris 33 -2.18 4.32 0.05 101.41 1 2 19 9 50.26 3 0.71

a. all species tended to shift earlier; F ratios and corresponding P values test whether the tendency to shift earlier was significantly different from zero.
b. 1 = egg; 2 = larva; 3 = pupa; 4 = adult.


TABLE A2. Summary of top models identified as having strong empirical support (Δ AICc 0–2) from the model selection approach; the relative importance of individual parameters among the top model subset is indicated.

Model
no. (
i)
Model
wt. (wi)
First
app.
b
Overw.
stage
No.
plants
Per.
nat.
Lat.
ext.
No. plants
 × 
per. nat.
No. plants
 × 
lat. ext.
Per. nat.
 × 
lat. ext.
1 0.394    
2 0.250  
3 0.144  
4 0.111  
5 0.101    
                   
w+ia 1.000
(cum. w+)
0.505 1.000 1.000 1.000 1.000 0.245 0.755 1.000

a. The relative importance value (w+ i) is the sum of the Akaike model weight (wi) for each model, i, the term occurs in. Relative importance values range from 0 to 1, with values closer to 1 being more important in explaining the degree of phenological advancement.
b. Abbreviated column names: First app. (Baseline annual date of first appearance in 1975); Overw. Stage (Overwintering stage); No. Plants (Number larval host plant species); Per. Nat. (Percent national 10km grid cells occupied); Lat. Ext. (UK latitudinal extent).


TABLE A3.  Blomberg's K statistic and associated P values based on the variance of phylogenetically independent contrasts relative to tip shuffling randomization (Blomberg and Garland 2002) for all species' traits identified during the model selection process, including the response, change per decade in date of first appearance.

Trait

K

P

Change per decade in date of first appearance

0.201

0.994

Number of larval host plant species

0.563

0.002

Overwintering stagea

0.541

0.001

Percent national 10-km grid cells occupied

0.498

0.028

Latitudinal extent

0.414

0.344

Baseline annual date of first appearance (1975)

0.442

0.003

a. Here overwintering stage is treated as continuous for analytical purposes, with the egg stage assigned as 1, larval stage as 2, pupal stage as 3, and adult stage as 4.

LITERATURE CITED

Blomberg, S. P., and T. Garland. 2002. Tempo and mode in evolution: phylogenetic inertia, adaptation and comparative methods. Journal of Evolutionary Biology 15:899–910.

Velleman, P. F., and R. E. Welsch. 1981. Efficient computing of regression diagnostics.


[Back to E092-083]