Bright Moonlight Triggers Natal Dispersal Departures

Bright moonlight triggers natal dispersal departures

Behavioral Ecology and Sociobiology

Vincenzo Penteriania,e, María del Mar Delgadob, Anna Kuparinenc,d, Pertti Saurolae, Jari Valkamae, Eino Salof, Jere Toivolag, Adrian Aebischerh and RaphaëlArlettazh

a Department of Conservation Biology, Estación Biológica de Doñana, C.S.I.C., c/ Americo Vespucio s/n, 41092 Seville, Spain (email: )

bMetapopulation Research Group, Department of Biosciences, University of FI-00014 Helsinki, Finland

c Department of Environmental Sciences, d Department of Biosciences, University of Helsinki, FI-00014 Helsinki, Finland

eFinnish Museum of Natural History, Zoology Unit, University of Helsinki, FI-00014 Helsinki, Finland

f Paloniityntie 90, Forssa, Finland

g Taatilantie 74, Tarttila, Finland

hDivision of Conservation Biology, Institute of Ecology and Evolution, University of Bern, Baltzerstrasse 6, 3012 Bern, Switzerland

Supplementary Material

Here we give details of linear mixed-effects model showing the effect of moon phase, age and sex of individuals on the decision to start dispersal:maximum models before simplification and steps of the optimal model selection are described.

Linear mixed-effects model fitted by maximum likelihood

Date of dispersal (juldisp)

Selection of random factor

reg1=gls(juldisp~z.age+sex+I(cos(z.rad))+I(sin(z.rad))+I(cos(2*z.rad))+I(sin(2*z.rad))+I(cos(z.rad)):z.age+I(sin(z.rad)):z.age + I(cos(2*z.rad)):z.age+I(sin(2*z.rad)):z.age, data=owls, method="REML")

reg2=lme(juldisp~z.age+sex+I(cos(z.rad))+I(sin(z.rad))+I(cos(2*z.rad))+I(sin(2*z.rad))+I(cos(z.rad)):z.age+I(sin(z.rad)):z.age + I(cos(2*z.rad)):z.age+I(sin(2*z.rad)):z.age, random=~1|year,data=owls, method="REML")

reg3=lme(juldisp~z.age+sex+I(cos(z.rad))+I(sin(z.rad))+I(cos(2*z.rad))+I(sin(2*z.rad))+I(cos(z.rad)):z.age+I(sin(z.rad)):z.age + I(cos(2*z.rad)):z.age+I(sin(2*z.rad)):z.age, random=~1|country,data=owls, method="REML")

reg4=lme(juldisp~z.age+sex+I(cos(z.rad))+I(sin(z.rad))+I(cos(2*z.rad))+I(sin(2*z.rad))+I(cos(z.rad)):z.age+I(sin(z.rad)):z.age + I(cos(2*z.rad)):z.age+I(sin(2*z.rad)):z.age, random=~1|nest,data=owls, method="REML")

reg5=lme(juldisp~z.age+sex+I(cos(z.rad))+I(sin(z.rad))+I(cos(2*z.rad))+I(sin(2*z.rad))+I(cos(z.rad)):z.age+I(sin(z.rad)):z.age + I(cos(2*z.rad)):z.age+I(sin(2*z.rad)):z.age, random=~1|year/country,data=owls, method="REML")

reg6=lme(juldisp~z.age+sex+I(cos(z.rad))+I(sin(z.rad))+I(cos(2*z.rad))+I(sin(2*z.rad))+I(cos(z.rad)):z.age+I(sin(z.rad)):z.age + I(cos(2*z.rad)):z.age+I(sin(2*z.rad)):z.age, random=~1|country/nest,data=owls, method="REML")

reg7=lme(juldisp~z.age+sex+I(cos(z.rad))+I(sin(z.rad))+I(cos(2*z.rad))+I(sin(2*z.rad))+I(cos(z.rad)):z.age+I(sin(z.rad)):z.age + I(cos(2*z.rad)):z.age+I(sin(2*z.rad)):z.age, random=~1|year/nest,data=owls, method="REML")

reg8=lme(juldisp~z.age+sex+I(cos(z.rad))+I(sin(z.rad))+I(cos(2*z.rad))+I(sin(2*z.rad))+I(cos(z.rad)):z.age+I(sin(z.rad)):z.age + I(cos(2*z.rad)):z.age+I(sin(2*z.rad)):z.age,random=~1|year/country/nest, data=owls, method="REML")

AIC(reg1,reg2,reg3,reg4,reg5,reg6,reg7,reg8)

df AIC

reg1 12 1351.641

reg2 13 1307.194

reg3 13 1222.852

reg4 13 1266.246

reg5 14 1249.273

reg6 14 1197.585

reg7 14 1095.048

reg8 15 1072.457

Full model

AIC BIC logLik

1100.168 1114.611 -535.084

Random effects:

Formula: ~1 | year

(Intercept)

SD: 0.00445135

Formula: ~1 | country inyear

(Intercept)

SD: 25.25119

Formula: ~1 | nest in country in year

(Intercept) Residual

SD: 20.14765 2.772645

Fixed effects: juldisp ~ z.age + sex + I(cos(z.rad)) + I(sin(z.rad)) + I(cos(2 * z.rad)) + I(sin(2 * z.rad)) + I(cos(z.rad)):z.age + I(sin(z.rad)):z.age + I(cos(2*z.rad)):z.age + I(sin(2*z.rad)):z.age

Value SE df t P

(Intercept) 266.15542 9.410758 65 28.282037 <0.001

z.age 21.29896 4.528525 56 4.703289 0.001

sex2 -1.88626 0.794091 56 -2.375369 0.0210

I(cos(z.rad)) -4.84824 7.820690 56 -0.619925 0.5378

I(sin(z.rad)) -1.74149 1.103271 56 -1.578475 0.1201

I(cos(2*z.rad)) 1.47196 3.456771 56 0.425820 0.6719

I(sin(2*z.rad)) 1.97532 1.699030 56 1.162614 0.2499

z.age:I(cos(z.rad))-4.23522 7.268408 56 -0.582689 0.5624

z.age:I(sin(z.rad))-0.78899 1.180648 56 -0.668266 0.5067

z.age:I(cos(2*z.rad))1.14209 3.147873 56 0.362812 0.7181

z.age:I(sin(2*z.rad))-0.93386 1.685271 56 -0.554129 0.5817

Model reduction steps

reg1=update(reg,~.-I(sin(2 * z.rad)))

anova(reg,reg1)