ELECTRONIC SUPPLEMENTARY MATERIALS
Do birds see the forest for the trees? Scale-dependent effects of tree
diversity on avian predation of artificial larvae
Evalyne W. Muiruri1, Kalle Rainio2, Julia Koricheva1
1
School of Biological Sciences, Royal Holloway University of London, Egham, Surrey,
TW20 0EX, United Kingdom
2
Department of Biology, University of Turku, FIN-20014 Turku, Finland
Correspondence: Evalyne W. Muiruri, School of Biological Sciences, Royal Holloway
University of London, Egham, Surrey, TW20 0EX, United Kingdom; Tel: +447807171932;
e-mail: evalyne.muiruri.2012@live.rhul.ac.uk
Online Resource 1 - Insect herbivory monitoring
Fifty leaves were sampled from two branches facing opposite directions from the lower canopy,
recording the number of exposed chewing insects (Agelastica alni L., weevils, Lepidoptera and
sawfly larvae) as well as the number of leaves exhibiting rolling, folding or mining damage by
concealed insect herbivores. As the number of exposed chewing insects was low and patchily
distributed, we used their presence or absence as a variable instead. For each leaf, chewing and
skeletonising damage was also scored in situ as follows: (1) 0.1-5% leaf area damaged, (2) 625% leaf area damaged, (3) 26-50% leaf area damaged, (4) 51-75% leaf area damaged and (5)
more than 75% of leaf area damaged. The number of leaves in each class was subsequently
multiplied by the mid-point of the category and the values summed to obtain an estimate of
herbivore damage per branch. These values, in turn, were averaged for each tree, generating
percentage values of leaf area damage per tree.
Online Resource 2 - Bird species, their life-history traits and number of territories observed during bird surveys at the Satakunta experiment. The
first survey (22nd May 2013) was conducted prior to data collection on insect herbivores or bird predation. The second survey (7th June 2013) was
conducted the day before the installation of artificial larvae and the third survey (12th June 2013) was completed during the predation experiment
on the 5th day after installation. Life trait data taken from (Barbaro and van Halder 2009) and references therein.
Bird species
No of territories/ Survey
Latin name
Common name
Foraging strategy
Breeding diet
Home range size
1st survey 2nd survey 3rd survey Total
Anthus trivialis
Tree pipit
ground gleaner
insects
medium (1–4 ha) 1
2
2
5
Carpodacus erythrinus Scarlet rosefinch
ground gleaner
seeds/ insects
small (<1 ha)
1
0
0
1
Carduelis spinus
Siskin
canopy gleaner
seeds/ insects
large (>5 ha)
2
0
0
2
Emberiza citrinella
Yellowhammer
ground gleaner
insects/ seeds
small (<1 ha)
2
2
0
4
Erithacus rubecula
Robin
understorey gleaner insects/ seeds
medium (1–4 ha) 1
2
5
8
Fringilla coelebs
Chaffinch
canopy gleaner
insects/ seeds
small (<1 ha)
4
5
5
14
Phylloscopus collybita
Chiffchaff
canopy gleaner
insects
small (<1 ha)
1
1
1
3
Parus cristatus
Crested tit
canopy gleaner
insects/ seeds
medium (1–4 ha) 1
0
1
2
Phylloscopus trochilus
Willow warbler
canopy gleaner
insects
small (<1 ha)
9
27
20
56
Parus major
Great tit
canopy gleaner
insects
small (<1 ha)
1
2
2
5
Prunella modularis
Dunnock
ground gleaner
insects
small (<1 ha)
4
0
2
6
Parus montanus
Willow tit
canopy gleaner
insects/ seeds
medium (1–4 ha) 0
1
0
1
Regulus regulus
Goldcrest
canopy gleaner
insects
medium (1–4 ha) 1
1
0
2
Sylvia atricapilla
Blackcap
understorey gleaner insects/ seeds
small (<1 ha)
0
4
0
4
Sylvia borin
Garden warbler
canopy gleaner
insects/ seeds
small (<1 ha)
2
3
3
8
Sylvia curruca
Lesser whitethroat
canopy gleaner
insects
small (<1 ha)
2
3
2
7
Scolopax rusticola
Woodcock
ground gleaner
worms/ insects medium (1–4 ha) 1
0
0
1
Turdus iliacus
Redwing
ground gleaner
worms/ insects small (<1 ha)
2
2
1
5
Turdus philomelos
Song thrush
ground prober
insects/ seeds
medium (1–4 ha) 1
1
4
6
Total No of territories
36
56
48
140
Total No of species
17
14
12
19
Online Resource 3 – Interactions between time, study area and selected variables. A summary of results from initial binomial GLMMs
including time (T) from installation of the artificial larvae, study area (A) and the predictor variables of interest in this experiment is presented
below. The four diversity variables; richness, pine density, alder density and birch density were tested at both plot and neighbourhood level
separately. In addition, we report statistics from models substituting diversity variables with either tree height variation within a plot or, tree
species identity of the focal tree.
Diversity variables
Richness
χ2 df
p
Plot
T
46.3 1 <0.001
A
11.3 1 <0.001
Variable
0.11 1
0.730
TxA
31.2 1 <0.001
T x Variable
0.11 1
0.735
A x Variable
0.91 1
0.340
T x A x Variable 2.93 1
0.087
Pine Density
χ2 df
p
45.4 1 <0.001
18.0 1 <0.001
19.1 1 <0.001
32.6 1 <0.001
0.02 1
0.882
0.02 1
0.896
0.09 1
0.770
Alder Density
χ2
p
45.1 1 <0.001
12.3 1 <0.001
1.09 1
0.296
32.8 1 <0.001
1.30 1
0.255
0.13 1
0.718
0.28 1
0.598
Birch Density
χ2 df
p
45.0 1 <0.001
15.2 1 <0.001
8.21 1
0.004
34.1 1 <0.001
1.48 1
0.223
0.07 1
0.795
0.01 1
0.919
Additional Variables
Tree Height Variation
χ2
df
p
45.0
1
<0.001
12.1
1
<0.001
0.12
1
0.727
30.7
1
<0.001
0.26
1
0.611
1.03
1
0.310
0.68
1
0.410
Richness
χ2 df
p
Neighbourhood
T
45.5 1 <0.001
A
10.8 1
0.001
Variable
2.08 1
0.150
TxA
32.1 1 <0.001
T x Variable
0.01 1
0.910
A x Variable
0.03 1
0.855
T x A x Variable 0.56 1
0.455
Pine Density
χ2 df
p
45.4 1 <0.001
18.4 1 <0.001
22.0 1 <0.001
32.9 1 <0.001
0.20 1
0.651
0.04 1
0.836
0.04 1
0.847
Alder Density
χ2 df
p
45.0 1 <0.001
12.8 1 <0.001
1.73 1
0.189
32.3 1 <0.001
1.57 1
0.210
1.02 1
0.311
0.26 1
0.610
Birch Density
χ2 df
p
44.8 1 <0.001
14.9 1 <0.001
5.97 1
0.015
34.5 1 <0.001
2.19 1
0.139
0.32 1
0.573
0.05 1
0.830
Tree Species Identity
χ2
df
p
45.7
2
<0.001
13.2
2
<0.001
21.5
2
<0.001
33.2
2
<0.001
1.58
2
0.453
2.10
2
0.350
3.14
2
0.208
Online Resource 4 – Effects of insect herbivore abundance and damage on bird predation
rates. The effects of (a) the presence of exposed chewing insects, b) the abundance of
concealed feeding insects or c) percentage of leaf area damaged by insects on the number of
damaged artificial larvae is shown for either birch or alder trees. The mean (±SE) number of
damaged larvae is shown in (a) and smoothed means drawn in panels (b) and (c) for each
species separately.